<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd" xmlns:googleplay="http://www.google.com/schemas/play-podcasts/1.0"><channel><title><![CDATA[Jared Heyman]]></title><description><![CDATA[Tech guy and investor. Founder of Rebel Fund and previously Pioneer Fund, CrowdMed (YC W13), Infosurv & Intengo (acq. LON: NFC). Ex-Bain consultant. Data nerd.]]></description><link>https://insights.rebelfund.vc</link><image><url>https://substackcdn.com/image/fetch/$s_!nWyU!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F03926e0f-3292-44e2-a47c-ab1a92cee09b_640x640.png</url><title>Jared Heyman</title><link>https://insights.rebelfund.vc</link></image><generator>Substack</generator><lastBuildDate>Sun, 22 Nov 2026 21:20:16 GMT</lastBuildDate><atom:link href="https://insights.rebelfund.vc/feed" rel="self" type="application/rss+xml"/><copyright><![CDATA[Jared Heyman]]></copyright><language><![CDATA[en]]></language><webMaster><![CDATA[jaredheyman@substack.com]]></webMaster><itunes:owner><itunes:email><![CDATA[jaredheyman@substack.com]]></itunes:email><itunes:name><![CDATA[Jared Heyman]]></itunes:name></itunes:owner><itunes:author><![CDATA[Jared Heyman]]></itunes:author><googleplay:owner><![CDATA[jaredheyman@substack.com]]></googleplay:owner><googleplay:email><![CDATA[jaredheyman@substack.com]]></googleplay:email><googleplay:author><![CDATA[Jared Heyman]]></googleplay:author><itunes:block><![CDATA[Yes]]></itunes:block><item><title><![CDATA[On why YC Series A rounds fell off a cliff]]></title><description><![CDATA[As my long-time readers know, Rebel Fund invests in the top 10% of Y Combinator startups each batch, now with a portfolio of 300+ incredible YC companies.]]></description><link>https://insights.rebelfund.vc/p/on-why-yc-series-a-rounds-fell-off-a-cliff-ab523211a2f2</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-why-yc-series-a-rounds-fell-off-a-cliff-ab523211a2f2</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Tue, 17 Nov 2026 19:15:13 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/52171ef4-e1c0-4871-ba24-e6152560bf4b_1024x495.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>As my long-time readers know, <a href="https://www.rebelfund.vc/">Rebel Fund</a> invests in the top 10% of <a href="https://www.ycombinator.com/">Y Combinator</a> startups each batch, now with a portfolio of 300+ incredible YC companies. Because our <a href="https://jaredheyman.medium.com/on-rebel-theorem-5-0-2c8270456e08">Rebel Theorem 5.0</a> ML/AL algorithm tracks hundreds of characteristics and outcomes of every YC company and founder in history, we often see shifts in YC startup performance in our data well before they show up in the headlines.</p><p>We noticed one particular shift that set off alarm bells&#8202;&#8212;&#8202;YC startup Series A rounds fell off a cliff starting about 5 years&nbsp;ago:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!_7RV!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97983535-20cf-420e-bcbd-faa8e42e50c0_1024x495.png 424w, https://substackcdn.com/image/fetch/$s_!_7RV!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97983535-20cf-420e-bcbd-faa8e42e50c0_1024x495.png 848w, https://substackcdn.com/image/fetch/$s_!_7RV!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97983535-20cf-420e-bcbd-faa8e42e50c0_1024x495.png 1272w, https://substackcdn.com/image/fetch/$s_!_7RV!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97983535-20cf-420e-bcbd-faa8e42e50c0_1024x495.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!_7RV!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97983535-20cf-420e-bcbd-faa8e42e50c0_1024x495.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/97983535-20cf-420e-bcbd-faa8e42e50c0_1024x495.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!_7RV!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97983535-20cf-420e-bcbd-faa8e42e50c0_1024x495.png 424w, https://substackcdn.com/image/fetch/$s_!_7RV!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97983535-20cf-420e-bcbd-faa8e42e50c0_1024x495.png 848w, https://substackcdn.com/image/fetch/$s_!_7RV!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97983535-20cf-420e-bcbd-faa8e42e50c0_1024x495.png 1272w, https://substackcdn.com/image/fetch/$s_!_7RV!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97983535-20cf-420e-bcbd-faa8e42e50c0_1024x495.png 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p>What happened? Did YC batches get too big? Is YC founder quality not what it used to be? Are macro factors at play? Today&#8217;s post sets out to answer&nbsp;this.</p><h3>Why the Series A rate&nbsp;matters</h3><p>Unsurprisingly, YC companies that raise a Series A have a dramatically higher survival rate than companies that don&#8217;t. According to our data, raising a Series A round cuts a company&#8217;s odds of shutting down by nearly 3x, and those that raise a Series C+ have a 10-year survival rate of&nbsp;&gt;95%.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!YKep!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F649c3b1a-7005-4401-ae95-5414fa237e05_1024x508.png 424w, https://substackcdn.com/image/fetch/$s_!YKep!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F649c3b1a-7005-4401-ae95-5414fa237e05_1024x508.png 848w, https://substackcdn.com/image/fetch/$s_!YKep!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F649c3b1a-7005-4401-ae95-5414fa237e05_1024x508.png 1272w, https://substackcdn.com/image/fetch/$s_!YKep!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F649c3b1a-7005-4401-ae95-5414fa237e05_1024x508.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!YKep!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F649c3b1a-7005-4401-ae95-5414fa237e05_1024x508.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/649c3b1a-7005-4401-ae95-5414fa237e05_1024x508.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!YKep!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F649c3b1a-7005-4401-ae95-5414fa237e05_1024x508.png 424w, https://substackcdn.com/image/fetch/$s_!YKep!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F649c3b1a-7005-4401-ae95-5414fa237e05_1024x508.png 848w, https://substackcdn.com/image/fetch/$s_!YKep!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F649c3b1a-7005-4401-ae95-5414fa237e05_1024x508.png 1272w, https://substackcdn.com/image/fetch/$s_!YKep!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F649c3b1a-7005-4401-ae95-5414fa237e05_1024x508.png 1456w" sizes="100vw"></picture><div></div></div></div></figure></div><h3>Is YC not what it used to&nbsp;be?</h3><p>There&#8217;s a perpetual meme amongst investors that &#8220;YC isn&#8217;t what it used to be&#8221;. I&#8217;ve been hearing this at every YC Demo Day since my own in 2013, but of course it&#8217;s no more true now than it was&nbsp;then.</p><p>If the 2021 and later batches were simply weaker, we&#8217;d expect to see two things in the data: 1) the slowdown would be confined to them, and 2) it would show up in our quality metrics of the founders and companies YC admitted. Neither turned out to be&nbsp;true.</p><h3>Every batch slowed down in the same&nbsp;year</h3><p>It turns out that Series A rounds slowed for <em>every</em> YC startup in 2023, regardless of whether they were admitted to YC one, two, or three plus years earlier. A drop in admissions quality around 2021 can&#8217;t reach back and slow down a company that went through YC in&nbsp;2017.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!DDe3!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F352b468a-0e0f-4620-9b7b-650e6b557c88_1024x390.png 424w, https://substackcdn.com/image/fetch/$s_!DDe3!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F352b468a-0e0f-4620-9b7b-650e6b557c88_1024x390.png 848w, https://substackcdn.com/image/fetch/$s_!DDe3!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F352b468a-0e0f-4620-9b7b-650e6b557c88_1024x390.png 1272w, https://substackcdn.com/image/fetch/$s_!DDe3!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F352b468a-0e0f-4620-9b7b-650e6b557c88_1024x390.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!DDe3!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F352b468a-0e0f-4620-9b7b-650e6b557c88_1024x390.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/352b468a-0e0f-4620-9b7b-650e6b557c88_1024x390.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!DDe3!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F352b468a-0e0f-4620-9b7b-650e6b557c88_1024x390.png 424w, https://substackcdn.com/image/fetch/$s_!DDe3!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F352b468a-0e0f-4620-9b7b-650e6b557c88_1024x390.png 848w, https://substackcdn.com/image/fetch/$s_!DDe3!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F352b468a-0e0f-4620-9b7b-650e6b557c88_1024x390.png 1272w, https://substackcdn.com/image/fetch/$s_!DDe3!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F352b468a-0e0f-4620-9b7b-650e6b557c88_1024x390.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><h3>The founders were just as&nbsp;strong</h3><p>At Rebel, we score every YC startup and founder across hundreds of dimensions using our proprietary <a href="https://jaredheyman.medium.com/on-rebel-theorem-5-0-2c8270456e08">Rebel Theorem 5.0</a> ML/AI model, looking at everything from founder education and work backgrounds, personality traits, <a href="https://jaredheyman.medium.com/on-the-5-archetypes-of-top-yc-founders-b0a81e8e09fc">archetypes</a>, entrepreneurial track record, <a href="https://jaredheyman.medium.com/on-founder-product-fit-7f1ad68d7af5">founder-product fit</a>, and&nbsp;more.</p><p>For this analysis, we used the same data to predict overall Series A rates by batch, and compared that to actual graduation rates. In the chart below, you&#8217;ll see that predicted vs actual rates held steady until the end of 2020, and then suddenly the actual rates fell far below what our model predicted:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!TbIo!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4b1d6fec-70f9-4545-81df-4fffe3c27d6e_1024x488.png 424w, https://substackcdn.com/image/fetch/$s_!TbIo!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4b1d6fec-70f9-4545-81df-4fffe3c27d6e_1024x488.png 848w, https://substackcdn.com/image/fetch/$s_!TbIo!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4b1d6fec-70f9-4545-81df-4fffe3c27d6e_1024x488.png 1272w, https://substackcdn.com/image/fetch/$s_!TbIo!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4b1d6fec-70f9-4545-81df-4fffe3c27d6e_1024x488.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!TbIo!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4b1d6fec-70f9-4545-81df-4fffe3c27d6e_1024x488.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/4b1d6fec-70f9-4545-81df-4fffe3c27d6e_1024x488.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!TbIo!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4b1d6fec-70f9-4545-81df-4fffe3c27d6e_1024x488.png 424w, https://substackcdn.com/image/fetch/$s_!TbIo!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4b1d6fec-70f9-4545-81df-4fffe3c27d6e_1024x488.png 848w, https://substackcdn.com/image/fetch/$s_!TbIo!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4b1d6fec-70f9-4545-81df-4fffe3c27d6e_1024x488.png 1272w, https://substackcdn.com/image/fetch/$s_!TbIo!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4b1d6fec-70f9-4545-81df-4fffe3c27d6e_1024x488.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>The slowdown also hit every tier of founder quality about&nbsp;equally.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!EiMG!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1491d49e-baaa-412e-a2fc-f50186f7f8e1_1024x506.png 424w, https://substackcdn.com/image/fetch/$s_!EiMG!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1491d49e-baaa-412e-a2fc-f50186f7f8e1_1024x506.png 848w, https://substackcdn.com/image/fetch/$s_!EiMG!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1491d49e-baaa-412e-a2fc-f50186f7f8e1_1024x506.png 1272w, https://substackcdn.com/image/fetch/$s_!EiMG!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1491d49e-baaa-412e-a2fc-f50186f7f8e1_1024x506.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!EiMG!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1491d49e-baaa-412e-a2fc-f50186f7f8e1_1024x506.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/1491d49e-baaa-412e-a2fc-f50186f7f8e1_1024x506.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!EiMG!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1491d49e-baaa-412e-a2fc-f50186f7f8e1_1024x506.png 424w, https://substackcdn.com/image/fetch/$s_!EiMG!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1491d49e-baaa-412e-a2fc-f50186f7f8e1_1024x506.png 848w, https://substackcdn.com/image/fetch/$s_!EiMG!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1491d49e-baaa-412e-a2fc-f50186f7f8e1_1024x506.png 1272w, https://substackcdn.com/image/fetch/$s_!EiMG!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1491d49e-baaa-412e-a2fc-f50186f7f8e1_1024x506.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>To me, these are the most striking numbers in the whole analysis. A top-quintile founding team that went through YC in 2022 had worse odds of raising a Series A than a bottom-quintile team from&nbsp;2018.</p><h3>The market&nbsp;turned</h3><p>The sudden slowdown in YC startup Series A rates coincides with interest rates climbing through 2022 and 2023. The pain started at the late stage, but global early-stage funding in late 2022 was <a href="https://news.crunchbase.com/venture/global-vc-funding-pullback-q3-2022-monthly-recap/">down 39%</a> year over year. As Series B and growth investors stopped writing checks, Series A investors lost confidence that their companies could raise the next round and stopped investing as aggressively.</p><p>It wasn&#8217;t just YC startups. <a href="https://carta.com/data/linkedin-archive/">Carta counted</a> 808 Series A rounds in 2023, down about 50% from 1,612 the year&nbsp;before.</p><p>So there&#8217;s our answer: <strong>It was a macroeconomic phenomenon.</strong></p><h3>So what from&nbsp;here?</h3><p>Many of the companies still standing from the tough 2021 time period onwards, which have founding teams just as strong as any before them, haven&#8217;t raised a Series A simply because the window was closed when they&nbsp;knocked.</p><p>However, there are early signs of a thaw. In 2025, the share of young YC companies raising a first Series A recovered to 4.2% for those one year past Demo Day and 7.3% for those two years past, up from 1.5% and 3.0% at the bottom. We&#8217;re still at roughly half the pre-2021 norm, so I wouldn&#8217;t call it a full recovery yet, but things are trending in the right direction.</p><p>The more interesting question is whether thanks to AI, companies will even <em>need to</em> raise a Series A at the same rates as before. In Rebel&#8217;s most recent vintages, we&#8217;re seeing incredible levels of revenue growth, capital efficiency, and profitability. If these companies can continue to fuel their growth on cash flow alone, they may not even want to raise any post-seed capital.</p><p>On the other hand, we&#8217;re also seeing companies take on dramatically more ambitious technological challenges than before, which could help maintain their appetite for venture capital after all, but that&#8217;s a topic for another blog&nbsp;post.</p>]]></content:encoded></item><item><title><![CDATA[On how to invest in venture]]></title><description><![CDATA[I&#8217;ve been a full-time venture capitalist for nearly a decade and seeded hundreds of top Y Combinator startups at Rebel Fund, now valued collectively in the tens of billions.]]></description><link>https://insights.rebelfund.vc/p/on-how-to-invest-in-venture-f53073e5217b</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-how-to-invest-in-venture-f53073e5217b</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Wed, 16 Sep 2026 15:03:23 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/9e766768-9659-4346-9fd3-0379dee97cf2_1024x683.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>I&#8217;ve been a <a href="https://www.linkedin.com/in/jaredheyman/">full-time venture capitalist</a> for nearly a decade and seeded hundreds of top <a href="https://www.ycombinator.com/">Y Combinator</a> startups at <a href="https://www.rebelfund.vc/">Rebel Fund</a>, now valued collectively in the tens of billions. So I&#8217;m often asked by other investors about how to get into venture investing. Since technology companies now dominate the S&amp;P 500, there&#8217;s more interest than ever in getting into tomorrow&#8217;s trillion-dollar tech companies early, and for good reason&#8202;&#8212;&#8202;where else can you enjoy a 1000x+ return on a single good investment?</p><p>Many of the <a href="https://jaredheyman.medium.com/">dozens of blog posts</a> I&#8217;ve published over the years cover various aspects of this, but I&#8217;ve never consolidated my learnings into a single post until now. If you want to get into venture investing but aren&#8217;t sure where to start, you&#8217;re in the right&nbsp;place.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!wDiT!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7f19e8c6-8bf7-4302-a99e-ac98a2d17e6a_1024x683.png 424w, https://substackcdn.com/image/fetch/$s_!wDiT!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7f19e8c6-8bf7-4302-a99e-ac98a2d17e6a_1024x683.png 848w, https://substackcdn.com/image/fetch/$s_!wDiT!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7f19e8c6-8bf7-4302-a99e-ac98a2d17e6a_1024x683.png 1272w, https://substackcdn.com/image/fetch/$s_!wDiT!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7f19e8c6-8bf7-4302-a99e-ac98a2d17e6a_1024x683.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!wDiT!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7f19e8c6-8bf7-4302-a99e-ac98a2d17e6a_1024x683.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/7f19e8c6-8bf7-4302-a99e-ac98a2d17e6a_1024x683.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!wDiT!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7f19e8c6-8bf7-4302-a99e-ac98a2d17e6a_1024x683.png 424w, https://substackcdn.com/image/fetch/$s_!wDiT!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7f19e8c6-8bf7-4302-a99e-ac98a2d17e6a_1024x683.png 848w, https://substackcdn.com/image/fetch/$s_!wDiT!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7f19e8c6-8bf7-4302-a99e-ac98a2d17e6a_1024x683.png 1272w, https://substackcdn.com/image/fetch/$s_!wDiT!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7f19e8c6-8bf7-4302-a99e-ac98a2d17e6a_1024x683.png 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p><strong>Question #1&#8202;&#8212;&#8202;Should I invest in startups directly or through a&nbsp;fund?</strong></p><p>The answer depends on who you are, but for the vast majority, the answer is a fund (especially for early-stage investing). Four questions decide&nbsp;it:</p><p><em>&#8212; Can you build a portfolio of 50+ startups?</em></p><p>As I discussed in <a href="https://jaredheyman.medium.com/on-the-power-law-of-y-combinator-startups-19cfb39863d6">On the power law of Y Combinator startups</a>, given the outcomes distribution and risk of early-stage investing, Sharpe ratios don&#8217;t start to stabilize until you have a portfolio of at least 50&#8211;75 companies and risk-adjusted returns further improve as you add shots on goal. This is more companies than most investors realize.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!h8-H!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a052a56-7e1d-47d2-a801-10a421c285c6_1024x768.png 424w, https://substackcdn.com/image/fetch/$s_!h8-H!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a052a56-7e1d-47d2-a801-10a421c285c6_1024x768.png 848w, https://substackcdn.com/image/fetch/$s_!h8-H!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a052a56-7e1d-47d2-a801-10a421c285c6_1024x768.png 1272w, https://substackcdn.com/image/fetch/$s_!h8-H!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a052a56-7e1d-47d2-a801-10a421c285c6_1024x768.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!h8-H!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a052a56-7e1d-47d2-a801-10a421c285c6_1024x768.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/2a052a56-7e1d-47d2-a801-10a421c285c6_1024x768.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!h8-H!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a052a56-7e1d-47d2-a801-10a421c285c6_1024x768.png 424w, https://substackcdn.com/image/fetch/$s_!h8-H!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a052a56-7e1d-47d2-a801-10a421c285c6_1024x768.png 848w, https://substackcdn.com/image/fetch/$s_!h8-H!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a052a56-7e1d-47d2-a801-10a421c285c6_1024x768.png 1272w, https://substackcdn.com/image/fetch/$s_!h8-H!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2a052a56-7e1d-47d2-a801-10a421c285c6_1024x768.png 1456w" sizes="100vw"></picture><div></div></div></div><figcaption class="image-caption">Monte Carlo simulation histogram based on all YC startups (see <a href="https://jaredheyman.medium.com/on-the-power-law-of-y-combinator-startups-19cfb39863d6">On the power law of Y Combinator startups</a>)</figcaption></figure></div><p><em>&#8212; Are you a good&nbsp;picker?</em></p><p>In a world where the top ~5% of investments drive ~90% of returns, picking winners is a critically important skill. To complicate matters, at the earlier stages (until around Series B) the good companies look a lot like the bad ones. Our solution at Rebel is a <a href="https://jaredheyman.medium.com/on-rebel-theorem-5-0-2c8270456e08">sophisticated ML/AI model</a> trained on every YC company, founder and outcome in history, but if you don&#8217;t have millions of dollars to invest in data science, you at least need to have good instincts on what makes an exceptional company and&nbsp;founder.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!8nWN!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F87418e0c-9d7d-42f3-862a-807400f64af3_1024x420.jpeg 424w, https://substackcdn.com/image/fetch/$s_!8nWN!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F87418e0c-9d7d-42f3-862a-807400f64af3_1024x420.jpeg 848w, https://substackcdn.com/image/fetch/$s_!8nWN!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F87418e0c-9d7d-42f3-862a-807400f64af3_1024x420.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!8nWN!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F87418e0c-9d7d-42f3-862a-807400f64af3_1024x420.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!8nWN!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F87418e0c-9d7d-42f3-862a-807400f64af3_1024x420.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/87418e0c-9d7d-42f3-862a-807400f64af3_1024x420.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!8nWN!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F87418e0c-9d7d-42f3-862a-807400f64af3_1024x420.jpeg 424w, https://substackcdn.com/image/fetch/$s_!8nWN!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F87418e0c-9d7d-42f3-862a-807400f64af3_1024x420.jpeg 848w, https://substackcdn.com/image/fetch/$s_!8nWN!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F87418e0c-9d7d-42f3-862a-807400f64af3_1024x420.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!8nWN!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F87418e0c-9d7d-42f3-862a-807400f64af3_1024x420.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">All YC startups by latest inferred valuation (see <a href="https://jaredheyman.medium.com/on-the-power-law-of-y-combinator-startups-19cfb39863d6">On the power law of Y Combinator startups</a>)</figcaption></figure></div><p><em>&#8212; Will the best founders let you&nbsp;in?</em></p><p>This is the question many newbies fail to ask themselves. Since roughly 1 out of 20 investments will drive the vast majority of your returns, even <em>a little bit</em> of adverse selection (i.e., not getting into that 1 deal) is fatal. Our solution at Rebel is a partnership team of <a href="https://www.rebelfund.vc/#section-partners">famous YC unicorn founders</a> that new YC founders can&#8217;t say no to, along with a carefully built, stellar reputation within our investment ecosystem. Regardless of your team and strategy, you need to have a very good answer to this question.</p><p><em>&#8212; Can you help your portfolio companies grow?</em></p><p>This is more important in theory than practice, which makes it important in practice.</p><p>After nearly a decade of venture investing, I&#8217;ve learned that great founders don&#8217;t really need me and I can&#8217;t help the mediocre ones. The best I can do is nudge the good founders towards greatness. That said, since money is a commodity, all founders end up choosing their investors on value-add, so being able to help portfolio companies grow is the cost of admission.</p><p>If your answer to <em>any</em> of the questions above is &#8220;no&#8221; then you should invest in funds. In fact, even if your answer is &#8220;yes&#8221; to all of them, you should <em>still</em> invest in funds if their answer is a stronger yes. The only exception is if you&#8217;re investing for non-financial reasons like learning, fun, or giving&nbsp;back.</p><p>I&#8217;ll go further, and say that even if you ultimately want to invest directly into startups, you should still start by investing in funds. At Rebel, we have many &#8216;strategic&#8217; LPs motivated beyond financial returns. For&nbsp;example:</p><p><strong>Direct deal flow&#8202;</strong>&#8212;&#8202;Many early-stage funds can help their LPs invest directly into portfolio startups alongside them, either at the initial investment stage or growth-stage follow-ons. This is typically done via a pooled SPV (Special Purpose Vehicle) or for a large LP, a dedicated sidecar vehicle. The advantage of doing &#8220;directs via funds&#8221; is the selection and access problems have been solved for you. With the right early-stage fund partnerships, you can build an institutional-grade directs portfolio at a fraction of the risk of going in alone, and a fraction of the cost of investing in a growth&nbsp;fund.</p><p><strong>Education</strong>&#8202;&#8212;&#8202;Watching fund managers do their thing is a great way of learning, and LPs get a front row seat. You can learn how professional managers think about the macro environment, technology trends, portfolio construction, deal selection, access, reporting, etc.</p><p><strong>Community&#8202;</strong>&#8212;&#8202;At Rebel, we have hundreds of LPs and bring them together for various <a href="https://www.rebelfund.vc/#section-events">events</a> where they can learn directly from our leadership, portfolio founders, guest speakers, and each other (and have&nbsp;fun!)</p><p><strong>Question #2&#8202;&#8212;&#8202;Should I invest in early or late-stage startups?</strong></p><p>Since we invest exclusively in seed-stage YC startups, my perspective is admittedly biased, but I&#8217;m a big fan of the early stage. The obvious reason is the closer you get to company formation, the bigger the financial upside, but there are other reasons as&nbsp;well:</p><p><strong>Diversification</strong>&#8202;&#8212;&#8202;Minimum check sizes are much lower for early-stage startups than late-stage, which allows more diversification for a given investment budget.</p><p><strong>Cost&#8202;</strong>&#8212;&#8202;While <a href="https://jaredheyman.medium.com/on-why-you-should-invest-in-the-most-expensive-yc-startups-d5edcc9cb68">more expensive YC startups</a> typically do better at the seed stage, the valuation premium grows exponentially as companies mature.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!mDGr!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd85b4efe-1cfe-481a-b173-eaf3166c1960_1024x512.jpeg 424w, https://substackcdn.com/image/fetch/$s_!mDGr!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd85b4efe-1cfe-481a-b173-eaf3166c1960_1024x512.jpeg 848w, https://substackcdn.com/image/fetch/$s_!mDGr!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd85b4efe-1cfe-481a-b173-eaf3166c1960_1024x512.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!mDGr!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd85b4efe-1cfe-481a-b173-eaf3166c1960_1024x512.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!mDGr!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd85b4efe-1cfe-481a-b173-eaf3166c1960_1024x512.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/d85b4efe-1cfe-481a-b173-eaf3166c1960_1024x512.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!mDGr!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd85b4efe-1cfe-481a-b173-eaf3166c1960_1024x512.jpeg 424w, https://substackcdn.com/image/fetch/$s_!mDGr!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd85b4efe-1cfe-481a-b173-eaf3166c1960_1024x512.jpeg 848w, https://substackcdn.com/image/fetch/$s_!mDGr!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd85b4efe-1cfe-481a-b173-eaf3166c1960_1024x512.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!mDGr!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd85b4efe-1cfe-481a-b173-eaf3166c1960_1024x512.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">See <a href="https://jaredheyman.medium.com/on-why-you-should-invest-in-the-most-expensive-yc-startups-d5edcc9cb684">On why you should invest in the most expensive YC&nbsp;startups</a></figcaption></figure></div><p><strong>Access</strong>&#8202;&#8212;&#8202;The access issue also gets harder as companies grow. By the time companies reach the Series B stage, it&#8217;s fairly obvious who the winners are, so unless you supported them early or are well-known in the industry, you probably won&#8217;t get into the best&nbsp;rounds.</p><p><strong>Fun</strong>&#8202;&#8212;&#8202;I think this one is often under-appreciated. Supporting founders at the earliest stages when they actually need your money is a lot more fun than fighting for scraps in some hot late-stage round. You get to work directly with incredibly smart people at the height of their optimism.</p><p>Many investors view late-stage startups as less risky since they&#8217;ve already achieved product-market fit (PMF), but I think they&#8217;re just trading company risk for macro risk. When you invest at the early-stage, you don&#8217;t need to worry about the macro environment because their valuation will almost certainly grow in the years ahead if they do well, regardless of whether current valuations are high or low. When you invest at the late-stage, you&#8217;re heavily exposed to a near-term crash in valuations or a cold IPO&nbsp;market.</p><p><strong>Question #3&#8202;&#8212;&#8202;Should I invest in big or small&nbsp;funds?</strong></p><p>The first thing to understand is how fund size dictates what needs to happen for you to make&nbsp;money.</p><p>A $100M fund needs $300M of attributable exit proceeds to return 3x gross, but a $1B fund needs $3B. Even if we generously assume the $1B fund manager is just as skilled, motivated, and selective, the mathematics are unavoidable, as is the rarity of multi-billion dollar&nbsp;paydays.</p><p>The case for big funds is they can keep funding their winners round after round instead of getting diluted, which matters when the power law is this steep. Also, their outcomes are less volatile, albeit at the cost of mediocrity, which some LPs oddly&nbsp;prefer.</p><p>The downside of big funds is the managers must demand ownership that the best founders won&#8217;t give them and feel pressure to write far more checks than their team can support. Big fund managers are also more motivated by management fees than carry, which creates an obvious misalignment with&nbsp;LPs.</p><p>While dispersion among smaller funds is greater, <a href="https://www.institutionalinvestor.com/article/2cjotgawx7nn70st06k8w/portfolio/big-vc-funds-are-underperforming-smaller-ones-and-their-future-is-dim">the data shows</a> they outperform the big funds both at the high-end and on average, which says everything you need to&nbsp;know.</p><p>At Rebel, we built a strategy and our fund size was a natural consequence, as opposed to picking a fund size and then building a strategy around it. While they won&#8217;t readily admit it, most big fund managers did the&nbsp;latter.</p><p><strong>Question #4&#8202;&#8212;&#8202;What is my investing thesis?</strong></p><p>There are several thousand new venture-backed startups launching each year in the US alone and you can&#8217;t invest in them all. Nor would you want to, since the median startup outcome is terrible. Startup investing is about <a href="https://medium.com/p/on-hitting-the-bullseye-6295e99d2a89">hitting the bullseye</a>, and you need a specific investing thesis to hit it reliably. And no, an investing thesis isn&#8217;t &#8220;AI&#8221; or &#8220;robotics&#8221;.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!07jc!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F11bc511f-5aa6-4962-bf7f-6f1fb45b1bec_1024x1031.jpeg 424w, https://substackcdn.com/image/fetch/$s_!07jc!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F11bc511f-5aa6-4962-bf7f-6f1fb45b1bec_1024x1031.jpeg 848w, https://substackcdn.com/image/fetch/$s_!07jc!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F11bc511f-5aa6-4962-bf7f-6f1fb45b1bec_1024x1031.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!07jc!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F11bc511f-5aa6-4962-bf7f-6f1fb45b1bec_1024x1031.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!07jc!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F11bc511f-5aa6-4962-bf7f-6f1fb45b1bec_1024x1031.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/11bc511f-5aa6-4962-bf7f-6f1fb45b1bec_1024x1031.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!07jc!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F11bc511f-5aa6-4962-bf7f-6f1fb45b1bec_1024x1031.jpeg 424w, https://substackcdn.com/image/fetch/$s_!07jc!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F11bc511f-5aa6-4962-bf7f-6f1fb45b1bec_1024x1031.jpeg 848w, https://substackcdn.com/image/fetch/$s_!07jc!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F11bc511f-5aa6-4962-bf7f-6f1fb45b1bec_1024x1031.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!07jc!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F11bc511f-5aa6-4962-bf7f-6f1fb45b1bec_1024x1031.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">Diversified YC Seed Strategy (see <a href="https://jaredheyman.medium.com/on-hitting-the-bullseye-6295e99d2a89">On hitting the bullseye</a>)</figcaption></figure></div><p>At Rebel, we target the top 10% of new YC startups, which we&#8217;re uniquely positioned to do with our partnership team, history, reputation, algorithm, etc. Just as great startups have <a href="https://jaredheyman.medium.com/on-founder-product-fit-7f1ad68d7af5">founder-product fit</a>, great investors have manager-thesis fit. I&#8217;ve seen many theses work, but only when there&#8217;s alignment between the managers and the fund&#8217;s unique investing strategy.</p><p>Even if you&#8217;re investing exclusively into funds, you still need a thesis. For example, one of our LPs only invests in funds that have unique access to a vetted ecosystem, like Y Combinator or the PayPal mafia. Another LP only invests in funds that can provide high-quality deal flow for their direct investments, and another only in data-driven funds. It&#8217;s not that other types of funds can&#8217;t achieve strong financial performance for these LPs, it&#8217;s that they aren&#8217;t in a good position to select and access&nbsp;them.</p><p>I&#8217;ll wrap up this post here. Sorry that it wasn&#8217;t as hard-hitting and data-driven as my other posts, but given the high risk and even higher return potential of venture investing, I had to put it out&nbsp;there.</p>]]></content:encoded></item><item><title><![CDATA[On why to ignore TechCrunch at YC Demo Day]]></title><description><![CDATA[I&#8217;ve always found it curious that TechCrunch publishes articles around every Y Combinator Demo Day with titles like &#8220;16 of the most interesting startups from YC W26 Demo Day&#8221;.]]></description><link>https://insights.rebelfund.vc/p/on-why-to-ignore-techcrunch-at-yc-demo-day-9f49d8b3d180</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-why-to-ignore-techcrunch-at-yc-demo-day-9f49d8b3d180</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Thu, 10 Sep 2026 14:33:32 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/6829965a-b23e-4bb8-a408-fe0b22e56aa2_836x394.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>I&#8217;ve always found it curious that <a href="https://techcrunch.com/">TechCrunch</a> publishes articles around every <a href="https://www.ycombinator.com/">Y Combinator</a> Demo Day with titles like &#8220;<a href="https://techcrunch.com/2026/03/26/16-of-the-most-interesting-startups-from-yc-w26-demo-day/">16 of the most interesting startups from YC W26 Demo Day</a>&#8221;. Apparently thousands of investors read these articles, but are journalists really in a position to tell them which YC startups will become tomorrow&#8217;s unicorns?</p><p>Since <a href="https://www.rebelfund.vc/">Rebel Fund</a> tracks millions of datapoints across every YC startup, founder, and outcome in history, in large part to train our proprietary <a href="https://jaredheyman.medium.com/on-rebel-theorem-5-0-2c8270456e08">Rebel Theorem 5.0</a> ML/AI algorithm, we&#8217;re in a great position to answer questions like this. So, I asked my data science team to analyze all 2,498 YC startups from the W21 to F24 batches vs the 194 startups that TechCrunch highlighted across its Demo Day articles to see whether these companies were more likely to raise a Series A or later (or fail) vs their YC peers. The short answer is that investors should ignore these TechCrunch articles.</p><ul><li><p><strong>The TechCrunch-highlighted startups performed the same as their YC batch mates.</strong> 18.0% reached Series A or later against a 17.6% cohort average, and 14.4% shut down against a 16.8% cohort average. Both gaps sit well inside what any random list of the same size would&nbsp;produce.</p></li><li><p><strong>They named winners in the same proportion that they named companies.</strong> TechCrunch highlighted 7.8% of the companies across those batches and 8.0% of those that reached Series A&#8202;&#8212;&#8202;exactly what you&#8217;d expect by pure&nbsp;chance.</p></li><li><p><strong>Most YC-focused venture funds don&#8217;t do any better.</strong> This was our most surprising finding and the one that will get me the most evil looks at Demo Day&nbsp;&#128563;</p></li></ul><p>Now here are the receipts:</p><p>This chart compares the Series A+ graduation rate and shutdown rate of TechCrunch&#8217;s picks vs their YC cohort. You&#8217;ll see TechCrunch&#8217;s companies perform only marginally better.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!8qBh!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F303ea05c-fb4c-45d0-a480-52923a40347c_836x394.jpeg 424w, https://substackcdn.com/image/fetch/$s_!8qBh!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F303ea05c-fb4c-45d0-a480-52923a40347c_836x394.jpeg 848w, https://substackcdn.com/image/fetch/$s_!8qBh!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F303ea05c-fb4c-45d0-a480-52923a40347c_836x394.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!8qBh!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F303ea05c-fb4c-45d0-a480-52923a40347c_836x394.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!8qBh!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F303ea05c-fb4c-45d0-a480-52923a40347c_836x394.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/303ea05c-fb4c-45d0-a480-52923a40347c_836x394.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!8qBh!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F303ea05c-fb4c-45d0-a480-52923a40347c_836x394.jpeg 424w, https://substackcdn.com/image/fetch/$s_!8qBh!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F303ea05c-fb4c-45d0-a480-52923a40347c_836x394.jpeg 848w, https://substackcdn.com/image/fetch/$s_!8qBh!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F303ea05c-fb4c-45d0-a480-52923a40347c_836x394.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!8qBh!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F303ea05c-fb4c-45d0-a480-52923a40347c_836x394.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p>To see how much of TechCrunch&#8217;s small &#8216;outperformance&#8217; is due to chance, we ran 20,000 Monte Carlo simulations to see where their predictions fall on the distribution curve, and they&#8217;re well within what&#8217;s considered random&nbsp;noise.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!nPvJ!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38e1c94d-c972-47b9-9e6f-b9619cba7290_856x360.jpeg 424w, https://substackcdn.com/image/fetch/$s_!nPvJ!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38e1c94d-c972-47b9-9e6f-b9619cba7290_856x360.jpeg 848w, https://substackcdn.com/image/fetch/$s_!nPvJ!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38e1c94d-c972-47b9-9e6f-b9619cba7290_856x360.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!nPvJ!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38e1c94d-c972-47b9-9e6f-b9619cba7290_856x360.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!nPvJ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38e1c94d-c972-47b9-9e6f-b9619cba7290_856x360.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/38e1c94d-c972-47b9-9e6f-b9619cba7290_856x360.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!nPvJ!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38e1c94d-c972-47b9-9e6f-b9619cba7290_856x360.jpeg 424w, https://substackcdn.com/image/fetch/$s_!nPvJ!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38e1c94d-c972-47b9-9e6f-b9619cba7290_856x360.jpeg 848w, https://substackcdn.com/image/fetch/$s_!nPvJ!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38e1c94d-c972-47b9-9e6f-b9619cba7290_856x360.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!nPvJ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F38e1c94d-c972-47b9-9e6f-b9619cba7290_856x360.jpeg 1456w" sizes="100vw"></picture><div></div></div></div></figure></div><p>In some batches the TechCrunch picks did better than average and in some they did worse&#8202;&#8212;&#8202;further evidence that chance is mostly at play&nbsp;here.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!kWzY!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc68ca547-063a-4854-91cd-18de6f3b51d9_904x450.jpeg 424w, https://substackcdn.com/image/fetch/$s_!kWzY!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc68ca547-063a-4854-91cd-18de6f3b51d9_904x450.jpeg 848w, https://substackcdn.com/image/fetch/$s_!kWzY!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc68ca547-063a-4854-91cd-18de6f3b51d9_904x450.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!kWzY!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc68ca547-063a-4854-91cd-18de6f3b51d9_904x450.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!kWzY!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc68ca547-063a-4854-91cd-18de6f3b51d9_904x450.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/c68ca547-063a-4854-91cd-18de6f3b51d9_904x450.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!kWzY!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc68ca547-063a-4854-91cd-18de6f3b51d9_904x450.jpeg 424w, https://substackcdn.com/image/fetch/$s_!kWzY!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc68ca547-063a-4854-91cd-18de6f3b51d9_904x450.jpeg 848w, https://substackcdn.com/image/fetch/$s_!kWzY!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc68ca547-063a-4854-91cd-18de6f3b51d9_904x450.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!kWzY!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc68ca547-063a-4854-91cd-18de6f3b51d9_904x450.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>When we looked at the mean return multiple of the TechCrunch-highlighted companies vs their peers, the numbers are nearly identical, again pointing more towards noise over&nbsp;signal.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!vmZu!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1cac7676-b9ba-4d8e-ac8b-9d5b223361fc_545x415.png 424w, https://substackcdn.com/image/fetch/$s_!vmZu!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1cac7676-b9ba-4d8e-ac8b-9d5b223361fc_545x415.png 848w, https://substackcdn.com/image/fetch/$s_!vmZu!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1cac7676-b9ba-4d8e-ac8b-9d5b223361fc_545x415.png 1272w, https://substackcdn.com/image/fetch/$s_!vmZu!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1cac7676-b9ba-4d8e-ac8b-9d5b223361fc_545x415.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!vmZu!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1cac7676-b9ba-4d8e-ac8b-9d5b223361fc_545x415.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/1cac7676-b9ba-4d8e-ac8b-9d5b223361fc_545x415.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!vmZu!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1cac7676-b9ba-4d8e-ac8b-9d5b223361fc_545x415.png 424w, https://substackcdn.com/image/fetch/$s_!vmZu!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1cac7676-b9ba-4d8e-ac8b-9d5b223361fc_545x415.png 848w, https://substackcdn.com/image/fetch/$s_!vmZu!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1cac7676-b9ba-4d8e-ac8b-9d5b223361fc_545x415.png 1272w, https://substackcdn.com/image/fetch/$s_!vmZu!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1cac7676-b9ba-4d8e-ac8b-9d5b223361fc_545x415.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>While we were at it, we compared the early success (Series A+ graduation) rate to failure (shutdown) rate of the TechCrunch &#8216;portfolio&#8217; vs the 7 most active venture funds investing into Y Combinator startups. In fairness to our journalist friends, most of these funds don&#8217;t perform any&nbsp;better.</p><p>This chart compares the portfolio early success-to-failure ratio of the YC-focused venture funds (including Rebel) vs TechCrunch. To protect the innocent, we redacted every fund name but our&nbsp;own.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!vo0h!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c18f012-a1a6-4bbe-a873-9b675827abee_1024x406.png 424w, https://substackcdn.com/image/fetch/$s_!vo0h!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c18f012-a1a6-4bbe-a873-9b675827abee_1024x406.png 848w, https://substackcdn.com/image/fetch/$s_!vo0h!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c18f012-a1a6-4bbe-a873-9b675827abee_1024x406.png 1272w, https://substackcdn.com/image/fetch/$s_!vo0h!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c18f012-a1a6-4bbe-a873-9b675827abee_1024x406.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!vo0h!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c18f012-a1a6-4bbe-a873-9b675827abee_1024x406.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/0c18f012-a1a6-4bbe-a873-9b675827abee_1024x406.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!vo0h!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c18f012-a1a6-4bbe-a873-9b675827abee_1024x406.png 424w, https://substackcdn.com/image/fetch/$s_!vo0h!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c18f012-a1a6-4bbe-a873-9b675827abee_1024x406.png 848w, https://substackcdn.com/image/fetch/$s_!vo0h!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c18f012-a1a6-4bbe-a873-9b675827abee_1024x406.png 1272w, https://substackcdn.com/image/fetch/$s_!vo0h!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0c18f012-a1a6-4bbe-a873-9b675827abee_1024x406.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>As you&#8217;ll see, 2 of the venture funds slightly underperformed TechCrunch, 2 slightly overperformed, and 3 did much better. Many of these funds have raised hundreds of millions of dollars. Rebel had the highest&nbsp;ratio.</p><p>To separate signal and noise, we analyzed which venture funds&#8217; performance was beyond what we&#8217;d expect due to chance at their portfolio size, and the same 3 funds came out on top. This means the other 4 funds, like TechCrunch, did about the same as throwing darts at a YC dartboard.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!uAhs!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F39774e47-3a3a-46c6-9a24-e38cb4375b17_996x481.png 424w, https://substackcdn.com/image/fetch/$s_!uAhs!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F39774e47-3a3a-46c6-9a24-e38cb4375b17_996x481.png 848w, https://substackcdn.com/image/fetch/$s_!uAhs!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F39774e47-3a3a-46c6-9a24-e38cb4375b17_996x481.png 1272w, https://substackcdn.com/image/fetch/$s_!uAhs!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F39774e47-3a3a-46c6-9a24-e38cb4375b17_996x481.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!uAhs!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F39774e47-3a3a-46c6-9a24-e38cb4375b17_996x481.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/39774e47-3a3a-46c6-9a24-e38cb4375b17_996x481.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!uAhs!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F39774e47-3a3a-46c6-9a24-e38cb4375b17_996x481.png 424w, https://substackcdn.com/image/fetch/$s_!uAhs!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F39774e47-3a3a-46c6-9a24-e38cb4375b17_996x481.png 848w, https://substackcdn.com/image/fetch/$s_!uAhs!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F39774e47-3a3a-46c6-9a24-e38cb4375b17_996x481.png 1272w, https://substackcdn.com/image/fetch/$s_!uAhs!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F39774e47-3a3a-46c6-9a24-e38cb4375b17_996x481.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Before I get tar and feathered for publishing this, a few disclaimers:</p><ul><li><p><strong>We did this analysis with publicly available data on YC startups and outcomes</strong> so anyone with the desire and resources can reproduce it.</p></li><li><p><strong>We only looked at early startup success vs failure metrics</strong> and not long-term performance (though the former is likely predictive of the&nbsp;latter).</p></li><li><p><strong>TechCrunch is not a venture fund</strong> and they do great work for the YC ecosystem as journalists by highlighting interesting startups.</p></li><li><p><strong>Investor performance at YC Demo Day averages is very good</strong> so even the &#8216;worst&#8217; funds in our analysis are probably making their LPs happy. <a href="https://www.mfmpod.com/videos/garry-tan-on-spotting-extreme-winners-advice-for-founders-in-ai-lessons-from-paul-graham/">Garry Tan said</a> that investors who backed YC companies from the 2018&#8211;2020 batches saw a median ~5x return, with the bottom quartile at&nbsp;~3x.</p></li></ul><p>So, the bottom line is TechCrunch&#8217;s picks don&#8217;t perform any better than chance, most YC-focused venture funds don&#8217;t pick much better, and only a few funds produce true alpha. Can anyone recommend a good bodyguard?</p>]]></content:encoded></item><item><title><![CDATA[On the YC S26 batch — Return of the Atoms]]></title><description><![CDATA[I&#8217;ve been to nearly every Y Combinator Demo Day since my own in 2013, so people often ask me what each new batch feels like in terms of trends and quality.]]></description><link>https://insights.rebelfund.vc/p/on-the-yc-s26-batch-return-of-the-atoms-2f6f47fd295a</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-the-yc-s26-batch-return-of-the-atoms-2f6f47fd295a</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Sat, 05 Sep 2026 14:34:23 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/ddd1bf48-4c85-4318-abd4-a8a87e17e274_1024x794.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>I&#8217;ve been to nearly every <a href="https://www.ycombinator.com/">Y Combinator</a> Demo Day since my own in 2013, so people often ask me what each new batch feels like in terms of trends and quality. Since <a href="https://www.rebelfund.vc/">Rebel Fund</a> invests in hundreds of <a href="https://jaredheyman.medium.com/on-why-you-should-invest-in-the-most-expensive-yc-startups-d5edcc9cb68">expensive YC startups</a> and maintains the world&#8217;s largest database of YC startups, founders, and outcomes outside of YC itself, I prefer to answer that question with data. So here we go for the S26&nbsp;batch!</p><p><strong>YC pivots to&nbsp;atoms</strong></p><p>Industrials (robotics, manufacturing, hardware, energy, defense, logistics) make up a whopping 23% of the 245 companies launched thus far in this batch vs a five-year average of just 6.6%. That&#8217;s the highest industrials share of any batch we&#8217;ve tracked and roughly double the last two batches, which were already trending&nbsp;upwards.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!PgLM!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F52e8aa6b-89e1-4151-b16b-d2dcc49b464b_1024x794.png 424w, https://substackcdn.com/image/fetch/$s_!PgLM!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F52e8aa6b-89e1-4151-b16b-d2dcc49b464b_1024x794.png 848w, https://substackcdn.com/image/fetch/$s_!PgLM!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F52e8aa6b-89e1-4151-b16b-d2dcc49b464b_1024x794.png 1272w, https://substackcdn.com/image/fetch/$s_!PgLM!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F52e8aa6b-89e1-4151-b16b-d2dcc49b464b_1024x794.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!PgLM!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F52e8aa6b-89e1-4151-b16b-d2dcc49b464b_1024x794.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/52e8aa6b-89e1-4151-b16b-d2dcc49b464b_1024x794.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!PgLM!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F52e8aa6b-89e1-4151-b16b-d2dcc49b464b_1024x794.png 424w, https://substackcdn.com/image/fetch/$s_!PgLM!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F52e8aa6b-89e1-4151-b16b-d2dcc49b464b_1024x794.png 848w, https://substackcdn.com/image/fetch/$s_!PgLM!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F52e8aa6b-89e1-4151-b16b-d2dcc49b464b_1024x794.png 1272w, https://substackcdn.com/image/fetch/$s_!PgLM!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F52e8aa6b-89e1-4151-b16b-d2dcc49b464b_1024x794.png 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p>The mirror image is SaaS, which fell to 12% of the batch vs a 28% five-year average. B2B software overall is down to 52%, its smallest share since&nbsp;S22.</p><p>In terms of themes, what I find most telling is that AI <em>appears</em> to be on a downward trend, with 67% of S26 companies carrying an AI tag, down from a peak of 83% in S25. Nobody tagged themselves as an &#8220;internet company&#8221; after 2005 because every company was one, and I think AI has quietly crossed the same threshold. It&#8217;s now the substrate rather than the pitch, and the more interesting question is what founders are pointing it at. In S26, the answer is often robots, factories, and&nbsp;defense.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!l3ld!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1e574417-86a7-41b8-aa8f-66c491ef96f8_1024x1185.png 424w, https://substackcdn.com/image/fetch/$s_!l3ld!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1e574417-86a7-41b8-aa8f-66c491ef96f8_1024x1185.png 848w, https://substackcdn.com/image/fetch/$s_!l3ld!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1e574417-86a7-41b8-aa8f-66c491ef96f8_1024x1185.png 1272w, https://substackcdn.com/image/fetch/$s_!l3ld!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1e574417-86a7-41b8-aa8f-66c491ef96f8_1024x1185.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!l3ld!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1e574417-86a7-41b8-aa8f-66c491ef96f8_1024x1185.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/1e574417-86a7-41b8-aa8f-66c491ef96f8_1024x1185.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!l3ld!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1e574417-86a7-41b8-aa8f-66c491ef96f8_1024x1185.png 424w, https://substackcdn.com/image/fetch/$s_!l3ld!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1e574417-86a7-41b8-aa8f-66c491ef96f8_1024x1185.png 848w, https://substackcdn.com/image/fetch/$s_!l3ld!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1e574417-86a7-41b8-aa8f-66c491ef96f8_1024x1185.png 1272w, https://substackcdn.com/image/fetch/$s_!l3ld!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1e574417-86a7-41b8-aa8f-66c491ef96f8_1024x1185.png 1456w" sizes="100vw"></picture><div></div></div></div></figure></div><p>YC has flirted with hard tech before. Sam Altman pushed it hard in the mid-2010s and produced a handful of real winners, but those companies were always a small minority of each batch. This wave is different in scale, and I believe in kind, because AI has collapsed the cost of the software layer that used to make hardware companies so slow and expensive to&nbsp;build.</p><p>Robotics founders in this batch are shipping control software in weeks that would have taken a team of ten over a year to ship pre-AI. I&#8217;ll also note, since it&#8217;s relevant to how Rebel invests, that <a href="https://jaredheyman.medium.com/on-rebel-theorem-5-0-2c8270456e08">Rebel Theorem</a> has historically found industrials companies to be more likely to succeed than consumer companies and much of software, not less. YC funds the best founders it can find regardless of sector, so I&#8217;ve always read its industry mix as a barometer of where the strongest technical people think the opportunity is. Right now, more than ever think it&#8217;s in things you can&nbsp;touch.</p><p><strong>The missing&nbsp;middle</strong></p><p>If you read my post <a href="https://jaredheyman.medium.com/on-the-new-y-combinator-3c28e548896c">On the &#8216;new&#8217; Y Combinator</a>, you know YC founders have gotten much younger under Garry Tan&#8217;s leadership. S26 pushes that to a new extreme, but in a way we haven&#8217;t exactly seen&nbsp;before.</p><p>Nearly 60% of S26 founders are under 25, up from 38% in the prior batch&#8202;&#8212;&#8202;the highest share in five years. The average age, though, barely moved. That&#8217;s because the growth in under-25s came almost entirely out of the 25&#8211;29 band, which fell to a five-year low of 23%, while founders in their 30s and 40s held&nbsp;steady.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Bx1d!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffd9a2f55-d98c-49d7-b651-f8c63a60c05c_1024x603.png 424w, https://substackcdn.com/image/fetch/$s_!Bx1d!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffd9a2f55-d98c-49d7-b651-f8c63a60c05c_1024x603.png 848w, https://substackcdn.com/image/fetch/$s_!Bx1d!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffd9a2f55-d98c-49d7-b651-f8c63a60c05c_1024x603.png 1272w, https://substackcdn.com/image/fetch/$s_!Bx1d!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffd9a2f55-d98c-49d7-b651-f8c63a60c05c_1024x603.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Bx1d!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffd9a2f55-d98c-49d7-b651-f8c63a60c05c_1024x603.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/fd9a2f55-d98c-49d7-b651-f8c63a60c05c_1024x603.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!Bx1d!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffd9a2f55-d98c-49d7-b651-f8c63a60c05c_1024x603.png 424w, https://substackcdn.com/image/fetch/$s_!Bx1d!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffd9a2f55-d98c-49d7-b651-f8c63a60c05c_1024x603.png 848w, https://substackcdn.com/image/fetch/$s_!Bx1d!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffd9a2f55-d98c-49d7-b651-f8c63a60c05c_1024x603.png 1272w, https://substackcdn.com/image/fetch/$s_!Bx1d!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffd9a2f55-d98c-49d7-b651-f8c63a60c05c_1024x603.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>That should give my long-time readers pause, because I&#8217;ve written that Rebel Theorem has identified a founder age &#8220;sweet spot&#8221; of 26&#8211;30 years, where the odds of a breakout outcome peak. Founders under 26 have historically been much more likely to end up with dead startups. S26 hollowed out exactly the cohort our model likes best, but I&#8217;m not panicked about it for two&nbsp;reasons.</p><p>First, as I&#8217;ve written before, founder age is one of hundreds of features Rebel Theorem weighs, and the interactions <em>between</em> them matter far more than any single one. Second, these young founders are unusually well matched to what they&#8217;re building. Education fit and work experience fit, key <a href="https://jaredheyman.medium.com/on-founder-product-fit-7f1ad68d7af5">founder-product fit</a> dimensions we score at Rebel, rose significantly.</p><p>However, other founder-product fit dimensions have fallen. Target-geography fit, for example, which measures how well a founder knows the market they&#8217;re selling into, fell significantly, as did relevant go-to-market experience.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!m8ba!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97791451-cc3e-42af-b202-233e4de0f3bf_1024x1355.png 424w, https://substackcdn.com/image/fetch/$s_!m8ba!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97791451-cc3e-42af-b202-233e4de0f3bf_1024x1355.png 848w, https://substackcdn.com/image/fetch/$s_!m8ba!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97791451-cc3e-42af-b202-233e4de0f3bf_1024x1355.png 1272w, https://substackcdn.com/image/fetch/$s_!m8ba!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97791451-cc3e-42af-b202-233e4de0f3bf_1024x1355.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!m8ba!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97791451-cc3e-42af-b202-233e4de0f3bf_1024x1355.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/97791451-cc3e-42af-b202-233e4de0f3bf_1024x1355.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!m8ba!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97791451-cc3e-42af-b202-233e4de0f3bf_1024x1355.png 424w, https://substackcdn.com/image/fetch/$s_!m8ba!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97791451-cc3e-42af-b202-233e4de0f3bf_1024x1355.png 848w, https://substackcdn.com/image/fetch/$s_!m8ba!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97791451-cc3e-42af-b202-233e4de0f3bf_1024x1355.png 1272w, https://substackcdn.com/image/fetch/$s_!m8ba!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97791451-cc3e-42af-b202-233e4de0f3bf_1024x1355.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Put the founder fit moves together and you get the S26 founder in one line: knows the technology unusually well, knows the customer unusually little. For a batch full of first-time founders selling to factories, hospitals, and the Pentagon, that&#8217;s a risk worth considering.</p><p><strong>Solo founders as a software&nbsp;factory</strong></p><p>The finding that surprised me the most: 19% of S26 companies have a single founder vs a five-year average of 8.1%. That&#8217;s the highest solo share in five years, and the third elevated batch in a row, despite YC&#8217;s historical aversion to solo founders.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!pYcL!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf15ba41-101f-4013-a2f3-1abc614ff2c1_1024x391.png 424w, https://substackcdn.com/image/fetch/$s_!pYcL!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf15ba41-101f-4013-a2f3-1abc614ff2c1_1024x391.png 848w, https://substackcdn.com/image/fetch/$s_!pYcL!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf15ba41-101f-4013-a2f3-1abc614ff2c1_1024x391.png 1272w, https://substackcdn.com/image/fetch/$s_!pYcL!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf15ba41-101f-4013-a2f3-1abc614ff2c1_1024x391.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!pYcL!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf15ba41-101f-4013-a2f3-1abc614ff2c1_1024x391.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/bf15ba41-101f-4013-a2f3-1abc614ff2c1_1024x391.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!pYcL!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf15ba41-101f-4013-a2f3-1abc614ff2c1_1024x391.png 424w, https://substackcdn.com/image/fetch/$s_!pYcL!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf15ba41-101f-4013-a2f3-1abc614ff2c1_1024x391.png 848w, https://substackcdn.com/image/fetch/$s_!pYcL!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf15ba41-101f-4013-a2f3-1abc614ff2c1_1024x391.png 1272w, https://substackcdn.com/image/fetch/$s_!pYcL!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf15ba41-101f-4013-a2f3-1abc614ff2c1_1024x391.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>I assumed this was the hard-tech wave again, i.e. a lone roboticist with a thesis project, but it&#8217;s the opposite. Solo founders in S26 are disproportionately building software (64% of them vs 52% for the batch overall). The solos are young as well, with 55% under 25, the same share as founders on two-person teams.</p><p>In my <a href="https://jaredheyman.medium.com/on-the-freakishly-strong-yc-w26-batch-056ccb666076">W26 post</a> I raised the question of whether Claude Code, Codex, and the like have changed the fundamentals of building a startup. I think this is the first hard evidence in our data that the answer is yes, at least for software. A 23-year-old first-time founder can now get a software company far enough alone to be worth YC&#8217;s money. Whether a solo founder can carry a hardware company is a separate question.</p><p>I&#8217;ve watched a lot of YC batches, and I can&#8217;t remember one this internally consistent: heavy on physical products, often built by very young technical founders out of university labs, many of which have never founded a company or raised a round. Whether they&#8217;ve picked markets that reward their technical expertise, or are about to learn how hard atoms and legacy industries really are, is the billion dollar question.</p>]]></content:encoded></item><item><title><![CDATA[On why you should invest in the most expensive YC startups]]></title><description><![CDATA[Before I proceed to enrage half of Silicon Valley with data, a quick introduction for the uninitiated:]]></description><link>https://insights.rebelfund.vc/p/on-why-you-should-invest-in-the-most-expensive-yc-startups-d5edcc9cb684</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-why-you-should-invest-in-the-most-expensive-yc-startups-d5edcc9cb684</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Thu, 27 Aug 2026 14:22:28 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/daeff4eb-7777-4fc4-ba95-e431546127b3_1024x438.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>Before I proceed to enrage half of Silicon Valley with data, a quick introduction for the uninitiated:</p><p><a href="https://www.rebelfund.vc/">Rebel Fund</a> invests in the top 10% of <a href="https://www.ycombinator.com/">Y Combinator</a> startups, with a portfolio of now 300+ companies valued collectively in the tens of billions. Along the way, we&#8217;ve invested millions into building the most comprehensive dataset of YC startups, founders, and outcomes that exists outside of YC itself, with millions of datapoints across every YC batch in history. We use this data primarily to train our proprietary <a href="https://jaredheyman.medium.com/on-rebel-theorem-5-0-2c8270456e08">Rebel Theorem 5.0</a> ML/AI model to predict YC startup success, but it&#8217;s also helped me publish <a href="https://jaredheyman.medium.com/">dozens of data-driven posts</a> answering interesting questions about YC startup investing. Today we&#8217;ll dive into whether the most expensive YC startups in each batch are worth the&nbsp;premium.</p><p>To answer this, we looked at 642 companies from the 2021&#8211;2024 batches where we have Demo Day SAFE valuation cap data. Here are our findings:</p><p><strong>Higher-priced companies do better. </strong>Companies priced above their batch median reached Series A or beyond 25% of the time, against 8% for those below it&#8202;&#8212;&#8202;that&#8217;s a <em><strong>three times higher</strong></em> Series A+ graduation rate for the more expensive companies.</p><p><strong>They also failed less. </strong>16% of below-median valuation companies have shut down vs 6% above the median&#8202;&#8212;&#8202;that&#8217;s a <em><strong>three times lower</strong></em> failure&nbsp;rate.</p><p><strong>Vintage does not explain it. </strong>Compare each company only against others from its own batch and the findings are consistent across vintages.</p><p><strong>Rebel&#8217;s involvement does not explain it.</strong> This holds across the companies we didn&#8217;t invest in (those we invest in perform better across valuations)</p><p><strong>Higher cap does not mean a bigger winner. </strong>This is the big caveat&#8202;&#8212;&#8202;we know that expensive companies enjoy higher Series A+ graduation rates and lower failure rates, but it&#8217;s unclear whether they&#8217;ll enjoy bigger&nbsp;exits.</p><p>Let me start by addressing the elephant in the room: YC startups are expensive and they&#8217;re only getting more&nbsp;so.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!d1tX!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97c6a44d-fe22-42d9-a282-2ab1a4f609bf_1024x438.jpeg 424w, https://substackcdn.com/image/fetch/$s_!d1tX!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97c6a44d-fe22-42d9-a282-2ab1a4f609bf_1024x438.jpeg 848w, https://substackcdn.com/image/fetch/$s_!d1tX!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97c6a44d-fe22-42d9-a282-2ab1a4f609bf_1024x438.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!d1tX!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97c6a44d-fe22-42d9-a282-2ab1a4f609bf_1024x438.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!d1tX!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97c6a44d-fe22-42d9-a282-2ab1a4f609bf_1024x438.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/97c6a44d-fe22-42d9-a282-2ab1a4f609bf_1024x438.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!d1tX!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97c6a44d-fe22-42d9-a282-2ab1a4f609bf_1024x438.jpeg 424w, https://substackcdn.com/image/fetch/$s_!d1tX!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97c6a44d-fe22-42d9-a282-2ab1a4f609bf_1024x438.jpeg 848w, https://substackcdn.com/image/fetch/$s_!d1tX!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97c6a44d-fe22-42d9-a282-2ab1a4f609bf_1024x438.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!d1tX!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F97c6a44d-fe22-42d9-a282-2ab1a4f609bf_1024x438.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>From the W21 to P26 batches, median YC SAFE caps of the companies Rebel interviewed nearly tripled from $15M to $40M. There is certainly a &#8216;YC premium&#8217; (though our outcomes data shows it&#8217;s worth&nbsp;paying).</p><p>Even within these already-expensive YC batches, there are a number of companies raising at valuation caps well above their batch average, giving pause to many seed investors. However, near-term outcomes are dramatically better for the more expensive companies&#8230;</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!bA-B!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff4d6df9b-1853-4bdd-8ffa-44cf029717b2_1024x488.jpeg 424w, https://substackcdn.com/image/fetch/$s_!bA-B!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff4d6df9b-1853-4bdd-8ffa-44cf029717b2_1024x488.jpeg 848w, https://substackcdn.com/image/fetch/$s_!bA-B!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff4d6df9b-1853-4bdd-8ffa-44cf029717b2_1024x488.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!bA-B!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff4d6df9b-1853-4bdd-8ffa-44cf029717b2_1024x488.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!bA-B!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff4d6df9b-1853-4bdd-8ffa-44cf029717b2_1024x488.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/f4d6df9b-1853-4bdd-8ffa-44cf029717b2_1024x488.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!bA-B!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff4d6df9b-1853-4bdd-8ffa-44cf029717b2_1024x488.jpeg 424w, https://substackcdn.com/image/fetch/$s_!bA-B!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff4d6df9b-1853-4bdd-8ffa-44cf029717b2_1024x488.jpeg 848w, https://substackcdn.com/image/fetch/$s_!bA-B!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff4d6df9b-1853-4bdd-8ffa-44cf029717b2_1024x488.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!bA-B!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff4d6df9b-1853-4bdd-8ffa-44cf029717b2_1024x488.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>&#8230;and this holds true at every price&nbsp;level&#8230;</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!xFPO!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F62bdbafb-dd34-4a7e-87cc-9f928421c5a9_1024x437.jpeg 424w, https://substackcdn.com/image/fetch/$s_!xFPO!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F62bdbafb-dd34-4a7e-87cc-9f928421c5a9_1024x437.jpeg 848w, https://substackcdn.com/image/fetch/$s_!xFPO!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F62bdbafb-dd34-4a7e-87cc-9f928421c5a9_1024x437.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!xFPO!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F62bdbafb-dd34-4a7e-87cc-9f928421c5a9_1024x437.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!xFPO!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F62bdbafb-dd34-4a7e-87cc-9f928421c5a9_1024x437.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/62bdbafb-dd34-4a7e-87cc-9f928421c5a9_1024x437.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!xFPO!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F62bdbafb-dd34-4a7e-87cc-9f928421c5a9_1024x437.jpeg 424w, https://substackcdn.com/image/fetch/$s_!xFPO!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F62bdbafb-dd34-4a7e-87cc-9f928421c5a9_1024x437.jpeg 848w, https://substackcdn.com/image/fetch/$s_!xFPO!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F62bdbafb-dd34-4a7e-87cc-9f928421c5a9_1024x437.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!xFPO!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F62bdbafb-dd34-4a7e-87cc-9f928421c5a9_1024x437.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>&#8230;and if you plot the relationship between SAFE caps at Demo Day vs whether companies go on to raise Series A+ rounds, you get a nearly perfect line, meaning the former is a very consistent predictor of the&nbsp;latter&#185;.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!yoLf!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F333f7ac8-24f9-4031-bd26-87c07375d052_1024x514.jpeg 424w, https://substackcdn.com/image/fetch/$s_!yoLf!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F333f7ac8-24f9-4031-bd26-87c07375d052_1024x514.jpeg 848w, https://substackcdn.com/image/fetch/$s_!yoLf!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F333f7ac8-24f9-4031-bd26-87c07375d052_1024x514.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!yoLf!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F333f7ac8-24f9-4031-bd26-87c07375d052_1024x514.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!yoLf!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F333f7ac8-24f9-4031-bd26-87c07375d052_1024x514.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/333f7ac8-24f9-4031-bd26-87c07375d052_1024x514.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!yoLf!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F333f7ac8-24f9-4031-bd26-87c07375d052_1024x514.jpeg 424w, https://substackcdn.com/image/fetch/$s_!yoLf!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F333f7ac8-24f9-4031-bd26-87c07375d052_1024x514.jpeg 848w, https://substackcdn.com/image/fetch/$s_!yoLf!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F333f7ac8-24f9-4031-bd26-87c07375d052_1024x514.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!yoLf!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F333f7ac8-24f9-4031-bd26-87c07375d052_1024x514.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Interestingly, Rebel portcos with lower Demo Day valuations perform nearly as well as those with higher valuations, and our companies perform much better than their YC peers across every valuation range.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!OFUH!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffed283ef-9b05-477a-a3c7-bcb0f01d92fa_1024x487.jpeg 424w, https://substackcdn.com/image/fetch/$s_!OFUH!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffed283ef-9b05-477a-a3c7-bcb0f01d92fa_1024x487.jpeg 848w, https://substackcdn.com/image/fetch/$s_!OFUH!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffed283ef-9b05-477a-a3c7-bcb0f01d92fa_1024x487.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!OFUH!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffed283ef-9b05-477a-a3c7-bcb0f01d92fa_1024x487.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!OFUH!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffed283ef-9b05-477a-a3c7-bcb0f01d92fa_1024x487.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/fed283ef-9b05-477a-a3c7-bcb0f01d92fa_1024x487.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!OFUH!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffed283ef-9b05-477a-a3c7-bcb0f01d92fa_1024x487.jpeg 424w, https://substackcdn.com/image/fetch/$s_!OFUH!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffed283ef-9b05-477a-a3c7-bcb0f01d92fa_1024x487.jpeg 848w, https://substackcdn.com/image/fetch/$s_!OFUH!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffed283ef-9b05-477a-a3c7-bcb0f01d92fa_1024x487.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!OFUH!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffed283ef-9b05-477a-a3c7-bcb0f01d92fa_1024x487.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>I think this is partially thanks to our portfolio support, but mostly a combination of our partners&#8217; picking ability and our <a href="https://jaredheyman.medium.com/on-rebel-theorem-5-0-2c8270456e08">Rebel Theorem</a> ML/AI algorithm, which looks at hundreds of characteristics of every YC startup and founder. The algorithm often picks up signals that are highly predictive of future outcomes yet unknown to other investors, and thus don&#8217;t drive up company valuations.</p><p>Still, Rebel tends to invest in the more expensive half of each YC batch, even though we don&#8217;t treat price as a signal one way or the&nbsp;other.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!pW3l!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf66e843-62c7-46ae-877b-62c3f312f46b_1024x348.jpeg 424w, https://substackcdn.com/image/fetch/$s_!pW3l!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf66e843-62c7-46ae-877b-62c3f312f46b_1024x348.jpeg 848w, https://substackcdn.com/image/fetch/$s_!pW3l!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf66e843-62c7-46ae-877b-62c3f312f46b_1024x348.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!pW3l!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf66e843-62c7-46ae-877b-62c3f312f46b_1024x348.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!pW3l!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf66e843-62c7-46ae-877b-62c3f312f46b_1024x348.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/af66e843-62c7-46ae-877b-62c3f312f46b_1024x348.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!pW3l!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf66e843-62c7-46ae-877b-62c3f312f46b_1024x348.jpeg 424w, https://substackcdn.com/image/fetch/$s_!pW3l!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf66e843-62c7-46ae-877b-62c3f312f46b_1024x348.jpeg 848w, https://substackcdn.com/image/fetch/$s_!pW3l!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf66e843-62c7-46ae-877b-62c3f312f46b_1024x348.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!pW3l!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faf66e843-62c7-46ae-877b-62c3f312f46b_1024x348.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>We also found the startups Rebel invests in enjoy a much higher success (Series A+ graduation) to failure (shut down) ratio than their YC peers. Our portfolio success-to-failure ratio is ~6:1 vs ~1:1 for the companies we passed&nbsp;on.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!IiAx!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3d6e5b14-f268-43ac-9c71-4f86658199b3_1024x526.jpeg 424w, https://substackcdn.com/image/fetch/$s_!IiAx!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3d6e5b14-f268-43ac-9c71-4f86658199b3_1024x526.jpeg 848w, https://substackcdn.com/image/fetch/$s_!IiAx!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3d6e5b14-f268-43ac-9c71-4f86658199b3_1024x526.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!IiAx!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3d6e5b14-f268-43ac-9c71-4f86658199b3_1024x526.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!IiAx!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3d6e5b14-f268-43ac-9c71-4f86658199b3_1024x526.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/3d6e5b14-f268-43ac-9c71-4f86658199b3_1024x526.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!IiAx!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3d6e5b14-f268-43ac-9c71-4f86658199b3_1024x526.jpeg 424w, https://substackcdn.com/image/fetch/$s_!IiAx!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3d6e5b14-f268-43ac-9c71-4f86658199b3_1024x526.jpeg 848w, https://substackcdn.com/image/fetch/$s_!IiAx!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3d6e5b14-f268-43ac-9c71-4f86658199b3_1024x526.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!IiAx!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3d6e5b14-f268-43ac-9c71-4f86658199b3_1024x526.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>And the stronger our interviewing partner&#8217;s recommendation was for a company, the better their&nbsp;outcome:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!TzMB!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F390fe0ee-561b-49ce-bf9b-cf67584ee098_1024x487.jpeg 424w, https://substackcdn.com/image/fetch/$s_!TzMB!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F390fe0ee-561b-49ce-bf9b-cf67584ee098_1024x487.jpeg 848w, https://substackcdn.com/image/fetch/$s_!TzMB!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F390fe0ee-561b-49ce-bf9b-cf67584ee098_1024x487.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!TzMB!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F390fe0ee-561b-49ce-bf9b-cf67584ee098_1024x487.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!TzMB!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F390fe0ee-561b-49ce-bf9b-cf67584ee098_1024x487.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/390fe0ee-561b-49ce-bf9b-cf67584ee098_1024x487.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!TzMB!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F390fe0ee-561b-49ce-bf9b-cf67584ee098_1024x487.jpeg 424w, https://substackcdn.com/image/fetch/$s_!TzMB!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F390fe0ee-561b-49ce-bf9b-cf67584ee098_1024x487.jpeg 848w, https://substackcdn.com/image/fetch/$s_!TzMB!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F390fe0ee-561b-49ce-bf9b-cf67584ee098_1024x487.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!TzMB!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F390fe0ee-561b-49ce-bf9b-cf67584ee098_1024x487.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>If you were to only invest in YC companies that 1) have a higher SAFE cap, and 2) Rebel invested in, you&#8217;d perform about <em><strong>6x better than YC averages</strong></em> in terms of Series A graduation rate. That would be a superb investment strategy, except very few of you could get into these rounds&nbsp;:-)</p><p>Before every investor who reads this starts aping into high-valuation YC startups and every YC founder starts hiking up their SAFE cap to improve their chances of raising a Series A, let me share some&nbsp;caveats:</p><ol><li><p><strong>Correlation does not imply causation.</strong> Near-term success is strongly <em>correlated</em> with higher valuation caps at Demo Day, but not <em>caused</em> by them. Companies that raise at higher valuations typically have stronger teams, bigger markets, and more traction to date. At Rebel, we often pass on high-valuation companies, even knowing these statistics, if we don&#8217;t feel their valuation is justified.</p></li><li><p><strong>Series A graduation rate isn&#8217;t destiny.</strong> While raising a Series A+ is definitely a good sign, what ultimately matters to founders and investors is exit valuation and dilution. This analysis shows that high-valuation companies are <em>safer</em>, but not necessarily <em>better </em>in terms of long-term outcomes.</p></li><li><p><strong>Buy low, sell high.</strong> The old investing adage still applies. The best skill a seed investor can have is identifying &#8216;hidden gem&#8217; startups with <em>lower</em> entry valuations that end up having exceptional exits. That&#8217;s what we aim to do at Rebel&#8202;&#8212;&#8202;we&#8217;re just not afraid to invest in a strong company that also happens to have a higher valuation, so long as it fits within our target&nbsp;range.</p></li></ol><p>Investors and founders, do as you wish with this analysis, please just don&#8217;t shoot the messenger &#128514;</p><p>&#185;The best-fit line has a log-odds coefficient of ~1.66, so doubling a company&#8217;s SAFE cap relative to its batch multiplies its Series A+ odds by ~3.1x. In other words, the more expensive companies are still &#8220;good buys&#8221; even after accounting for their higher&nbsp;prices.</p>]]></content:encoded></item><item><title><![CDATA[On Rebel Theorem 5.0]]></title><description><![CDATA[When Rebel Fund began our journey in 2019, our thesis was simple:]]></description><link>https://insights.rebelfund.vc/p/on-rebel-theorem-5-0-2c8270456e08</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-rebel-theorem-5-0-2c8270456e08</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Mon, 17 Aug 2026 13:23:16 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/15a5fde7-f0c4-4510-a56d-cec43a20db6f_1024x628.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>When <a href="https://www.rebelfund.vc/">Rebel Fund</a> began our journey in 2019, our thesis was&nbsp;simple:</p><ol><li><p><a href="https://www.ycombinator.com/">Y Combinator</a> startups are the best investment class in&nbsp;venture</p></li><li><p><a href="https://www.rebelfund.vc/#section-partners">YC unicorn founder partners </a>can best select/access the top YC&nbsp;startups</p></li><li><p>ML (+ later AI) models can best screen startups for partner interviews</p></li></ol><p>Over the past 7 years, all of these have played out. We&#8217;ve now seeded over 300 top YC startups, including several unicorns, and established a reputation as one of the most sought-after investors in the YC ecosystem.</p><p>We&#8217;ve also invested millions of dollars into developing the world&#8217;s most sophisticated ML/AI algorithm for predicting YC startup success at the pre-Demo Day stage. I&#8217;ve written about our various algorithm iterations over the years, most recently <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">Rebel Theorem 4.0</a> and our new <a href="https://jaredheyman.medium.com/on-what-actually-predicts-yc-startup-success-2b599c02537a">AI composite scores</a>, and now I&#8217;m pleased to announce Rebel Theorem&nbsp;5.0.</p><p>As with its predecessors, RT 5.0 is trained on millions of datapoints on every YC startup, founder, and outcome in history. We track and measure every company and team across hundreds of dimensions, and have identified the exact combination or characteristics (or &#8220;features&#8221; in ML speak) that most accurately predict YC startup&nbsp;success.</p><p>Given how much we&#8217;ve invested into developing RT 5.0 and how powerful it&#8217;s become, I&#8217;m going to be more tight-lipped about the inner workings of the algorithm than in my previous posts. Today&#8217;s post will focus more on how we validate and use the algorithm than how it works under the&nbsp;hood.</p><p>To start, let me explain what the latest algorithm actually does: <em><strong>predicts a company&#8217;s likelihood of enjoying a Series A&nbsp;markup</strong></em></p><p>A lot investors are surprised by that, because Series A rounds aren&#8217;t what drive portfolio returns for seed funds like us&#8202;&#8212;&#8202;it&#8217;s multi-billion dollar exits. However, we&#8217;ve found in backtesting that it&#8217;s impossible to predict at the seed stage exactly which startups will become unicorns, because unicorns are rare (around ~6% of YC startups) and there&#8217;s too much noise in the data to identify them pre-Demo Day, even with a model as robust as&nbsp;ours.</p><p>However, we <em><strong>can</strong></em> accurately predict a company&#8217;s likelihood or enjoying a Series A markup, and then put only the most likely winners in front of our Rebel partners for further diligence. Since ~20% of YC startups that achieve this milestone ultimately become a unicorn, that&#8217;s sufficient to enrich our portfolio with enough unicorns to be a top-decile seed&nbsp;fund.</p><h3>Rebel Theorem 5.0 model performance</h3><p>Now the billion dollar question: <em><strong>how well does the model&nbsp;work?</strong></em></p><p>Let&#8217;s start by comparing it to its predecessor, <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">Rebel Theorem 4.0</a> + our AI composite scores:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!k-l_!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8598baf-af04-44dc-933d-a0c8aa848848_1024x628.jpeg 424w, https://substackcdn.com/image/fetch/$s_!k-l_!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8598baf-af04-44dc-933d-a0c8aa848848_1024x628.jpeg 848w, https://substackcdn.com/image/fetch/$s_!k-l_!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8598baf-af04-44dc-933d-a0c8aa848848_1024x628.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!k-l_!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8598baf-af04-44dc-933d-a0c8aa848848_1024x628.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!k-l_!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8598baf-af04-44dc-933d-a0c8aa848848_1024x628.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/a8598baf-af04-44dc-933d-a0c8aa848848_1024x628.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!k-l_!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8598baf-af04-44dc-933d-a0c8aa848848_1024x628.jpeg 424w, https://substackcdn.com/image/fetch/$s_!k-l_!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8598baf-af04-44dc-933d-a0c8aa848848_1024x628.jpeg 848w, https://substackcdn.com/image/fetch/$s_!k-l_!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8598baf-af04-44dc-933d-a0c8aa848848_1024x628.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!k-l_!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa8598baf-af04-44dc-933d-a0c8aa848848_1024x628.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">RT 5.0 vs RT 4.0 (2023&#8211;2024 backtest)</figcaption></figure></div><p>This chart compares the composite prediction scores of RT 5.0 (blue) with RT 4.0 (orange) in terms of how many future winners were predicted per 250 companies interviewed, and how many &#8216;dead&#8217; companies slipped through. This backtest was based on 2023 &amp; 2024, which is why so many companies haven&#8217;t yet been declared winners or dead. As you&#8217;ll see, the newer model produced a modest ~7% improvement in predicting winners and a more dramatic ~55% reduction in false-positives.</p><p>More importantly though, the ~7% improvement in predicting winners is concentrated heavily on the mega-winners that really drive returns for seed funds like&nbsp;us.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!yvKJ!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F57ce7544-9686-4063-a6bd-e0becad25bd1_1024x702.png 424w, https://substackcdn.com/image/fetch/$s_!yvKJ!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F57ce7544-9686-4063-a6bd-e0becad25bd1_1024x702.png 848w, https://substackcdn.com/image/fetch/$s_!yvKJ!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F57ce7544-9686-4063-a6bd-e0becad25bd1_1024x702.png 1272w, https://substackcdn.com/image/fetch/$s_!yvKJ!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F57ce7544-9686-4063-a6bd-e0becad25bd1_1024x702.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!yvKJ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F57ce7544-9686-4063-a6bd-e0becad25bd1_1024x702.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/57ce7544-9686-4063-a6bd-e0becad25bd1_1024x702.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!yvKJ!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F57ce7544-9686-4063-a6bd-e0becad25bd1_1024x702.png 424w, https://substackcdn.com/image/fetch/$s_!yvKJ!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F57ce7544-9686-4063-a6bd-e0becad25bd1_1024x702.png 848w, https://substackcdn.com/image/fetch/$s_!yvKJ!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F57ce7544-9686-4063-a6bd-e0becad25bd1_1024x702.png 1272w, https://substackcdn.com/image/fetch/$s_!yvKJ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F57ce7544-9686-4063-a6bd-e0becad25bd1_1024x702.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>This chart below shows what our actual Rebel partner interview dockets would have looked like those years using the new model. As you&#8217;ll see, it heavily enriches our limited partner interview slots with future winners and wastes very little of our partners&#8217; time on companies that die (there are less winners thus far in the newer batches just because they&#8217;re still&nbsp;young).</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!E6Ph!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa71b3722-dde2-46cb-a217-474b508b631d_1024x534.jpeg 424w, https://substackcdn.com/image/fetch/$s_!E6Ph!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa71b3722-dde2-46cb-a217-474b508b631d_1024x534.jpeg 848w, https://substackcdn.com/image/fetch/$s_!E6Ph!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa71b3722-dde2-46cb-a217-474b508b631d_1024x534.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!E6Ph!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa71b3722-dde2-46cb-a217-474b508b631d_1024x534.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!E6Ph!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa71b3722-dde2-46cb-a217-474b508b631d_1024x534.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/a71b3722-dde2-46cb-a217-474b508b631d_1024x534.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!E6Ph!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa71b3722-dde2-46cb-a217-474b508b631d_1024x534.jpeg 424w, https://substackcdn.com/image/fetch/$s_!E6Ph!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa71b3722-dde2-46cb-a217-474b508b631d_1024x534.jpeg 848w, https://substackcdn.com/image/fetch/$s_!E6Ph!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa71b3722-dde2-46cb-a217-474b508b631d_1024x534.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!E6Ph!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa71b3722-dde2-46cb-a217-474b508b631d_1024x534.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">Rebel partner interview dockets backtest (2023&#8211;2024 backtest)</figcaption></figure></div><p>Pharmaceutical companies know a drug is working when a larger dose has a greater effect (dose-response curve). Likewise, we know our model is working when a higher success probability score correlates with actual better company outcomes.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!dhx0!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f9e81f2-ff56-419a-a04d-0c8eef2cc9e5_1024x646.jpeg 424w, https://substackcdn.com/image/fetch/$s_!dhx0!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f9e81f2-ff56-419a-a04d-0c8eef2cc9e5_1024x646.jpeg 848w, https://substackcdn.com/image/fetch/$s_!dhx0!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f9e81f2-ff56-419a-a04d-0c8eef2cc9e5_1024x646.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!dhx0!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f9e81f2-ff56-419a-a04d-0c8eef2cc9e5_1024x646.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!dhx0!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f9e81f2-ff56-419a-a04d-0c8eef2cc9e5_1024x646.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/3f9e81f2-ff56-419a-a04d-0c8eef2cc9e5_1024x646.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!dhx0!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f9e81f2-ff56-419a-a04d-0c8eef2cc9e5_1024x646.jpeg 424w, https://substackcdn.com/image/fetch/$s_!dhx0!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f9e81f2-ff56-419a-a04d-0c8eef2cc9e5_1024x646.jpeg 848w, https://substackcdn.com/image/fetch/$s_!dhx0!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f9e81f2-ff56-419a-a04d-0c8eef2cc9e5_1024x646.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!dhx0!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f9e81f2-ff56-419a-a04d-0c8eef2cc9e5_1024x646.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">RT 5.0 score ranks vs hit rate (2023&#8211;2024 backtest)</figcaption></figure></div><p>Here we can see that the higher a company&#8217;s RT 5.0 score, the more likely they were to actually enjoy a Series A markup. The top 10 companies in each batch had an extremely high hit rate of 42%&#8202;&#8212;&#8202;nearly 3x the batch&nbsp;average.</p><p>Of course this begs the question, why not just interview the top 10 companies as ranked by our model? The answer is that while interviewing more companies <em>decreases</em> our Series A hit rate, it <em>increases</em> our chances of hitting a super-outlier that can drive 100x+ returns for the fund. We&#8217;ve determined ~50 interviews per batch to be the optimal balance between risk (investing in too many losers) with reward (investing in enough mega-winners). See my post <a href="https://medium.com/p/on-hitting-the-bullseye-6295e99d2a89">On hitting the bullseye</a> to understand why a diversified portfolio is so important for seed investing.</p><p>Here&#8217;s the actual gain&nbsp;curve:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!AMHn!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faab23af6-2c7d-46f6-8ff7-08e0f07af754_1024x571.jpeg 424w, https://substackcdn.com/image/fetch/$s_!AMHn!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faab23af6-2c7d-46f6-8ff7-08e0f07af754_1024x571.jpeg 848w, https://substackcdn.com/image/fetch/$s_!AMHn!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faab23af6-2c7d-46f6-8ff7-08e0f07af754_1024x571.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!AMHn!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faab23af6-2c7d-46f6-8ff7-08e0f07af754_1024x571.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!AMHn!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faab23af6-2c7d-46f6-8ff7-08e0f07af754_1024x571.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/aab23af6-2c7d-46f6-8ff7-08e0f07af754_1024x571.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!AMHn!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faab23af6-2c7d-46f6-8ff7-08e0f07af754_1024x571.jpeg 424w, https://substackcdn.com/image/fetch/$s_!AMHn!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faab23af6-2c7d-46f6-8ff7-08e0f07af754_1024x571.jpeg 848w, https://substackcdn.com/image/fetch/$s_!AMHn!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faab23af6-2c7d-46f6-8ff7-08e0f07af754_1024x571.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!AMHn!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Faab23af6-2c7d-46f6-8ff7-08e0f07af754_1024x571.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">Cumulative gain (lift) curve for RT 5.0 (2023&#8211;2024 backtest)</figcaption></figure></div><p>Last but not least, I&#8217;ll share a chart that shows the correlation between RT 5.0 markup probability scores vs actual Series A&nbsp;markups.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!fTY6!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcd3af298-0d68-4ee0-81f9-0fd2ab356e6e_1024x637.jpeg 424w, https://substackcdn.com/image/fetch/$s_!fTY6!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcd3af298-0d68-4ee0-81f9-0fd2ab356e6e_1024x637.jpeg 848w, https://substackcdn.com/image/fetch/$s_!fTY6!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcd3af298-0d68-4ee0-81f9-0fd2ab356e6e_1024x637.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!fTY6!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcd3af298-0d68-4ee0-81f9-0fd2ab356e6e_1024x637.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!fTY6!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcd3af298-0d68-4ee0-81f9-0fd2ab356e6e_1024x637.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/cd3af298-0d68-4ee0-81f9-0fd2ab356e6e_1024x637.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!fTY6!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcd3af298-0d68-4ee0-81f9-0fd2ab356e6e_1024x637.jpeg 424w, https://substackcdn.com/image/fetch/$s_!fTY6!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcd3af298-0d68-4ee0-81f9-0fd2ab356e6e_1024x637.jpeg 848w, https://substackcdn.com/image/fetch/$s_!fTY6!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcd3af298-0d68-4ee0-81f9-0fd2ab356e6e_1024x637.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!fTY6!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcd3af298-0d68-4ee0-81f9-0fd2ab356e6e_1024x637.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">RT 5.0 markup probability scores vs actual markups (2023&#8211;2024 backtest)</figcaption></figure></div><p>As you&#8217;ll see, the model is extremely well-calibrated to its stated objective. When a company gets a markup probability score of ~30% then their chances of enjoying a Series A markup really is&nbsp;~30%.</p><p>I could share more charts and graphs, but the bottom line is our model works. You&#8217;d be hard-pressed to find another ML/AI model for predicting startup success that&#8217;s nearly as accurate and well-validated.</p><p>Ironically, the model doesn&#8217;t have to work as well as it does for us to maintain world-class fund performance&#8202;&#8212;&#8202;just not under-performing YC averages is enough. YC now has $1.3T in portfolio company value, well over 100 unicorns, a 3x unicorn rate vs non-YC, and the 13x TVPI (2014&#8211;2018 vintages) for Demo Day investors overall.</p><p>I know many of you are curious how Rebel Theorem 5.0 actually works, but unfortunately I can&#8217;t share anything beyond what&#8217;s in <a href="https://jaredheyman.medium.com/">my previous posts</a> about what actually predicts YC startup performance. In today&#8217;s environment, the stakes are just too high&nbsp;:-)</p>]]></content:encoded></item><item><title><![CDATA[On what actually predicts YC startup success]]></title><description><![CDATA[Since Rebel Fund invests exclusively in seed-stage Y Combinator startups, the dozens of blog posts I&#8217;ve published over the years are largely focused on what separates the future YC unicorns from the rest of the pack.]]></description><link>https://insights.rebelfund.vc/p/on-what-actually-predicts-yc-startup-success-2b599c02537a</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-what-actually-predicts-yc-startup-success-2b599c02537a</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Thu, 30 Apr 2026 23:55:18 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/b067105a-909d-4d82-bc70-0a63ff400232_1024x647.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>Since <a href="https://www.rebelfund.vc/">Rebel Fund</a> invests exclusively in seed-stage <a href="https://www.ycombinator.com/">Y Combinator</a> startups, the <a href="https://jaredheyman.medium.com/">dozens of blog posts</a> I&#8217;ve published over the years are largely focused on what separates the future YC unicorns from the rest of the pack. As my long-time readers know, Rebel maintains the most comprehensive database of YC startups, founders, and outcomes that exists outside of YC itself, and we use the millions of data points we&#8217;ve collected to train our proprietary Rebel Theorem ML algorithm to predict which YC startups are most likely to deliver outsized investor&nbsp;returns.</p><p>Last year I published <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">On Rebel Theorem 4.0</a>, unveiling the latest release of that algorithm after 5+ years of development. But our work doesn&#8217;t stop there&#8202;&#8212;&#8202;we also now compliment our core ML algorithm with a series of AI-derived scores on each startup to create a &#8220;composite&#8221; ML/AI score that is <em>much</em> more predictive of YC startup success than Rebel Theorem scores&nbsp;alone.</p><p>This post takes a step back from Rebel Theorem to focus on some of the AI-derived scores that we now assign to every YC startup before Demo Day, and which best predict a startup&#8217;s likelihood to raise a Series A round or&nbsp;beyond.</p><h3>The setup</h3><p>To answer the question rigorously, we worked from a clean dataset of 1,275 startups across five recent YC batches (S22 through S24). Of those companies, roughly 100&#8202;&#8212;&#8202;about 8%&#8202;&#8212;&#8202;have already raised a Series A round or beyond&#185;. Before I talk about which startup characteristics produce the highest statistical lift over that baseline, I&#8217;d like to illustrate just how powerful our composite ML/AI scores&nbsp;are.</p><p>The chart below shows how a company&#8217;s Series A+ likelihood climbs by our composite ML/AI score band. Companies in the highest score band raise a Series A or beyond <strong>at ~4x the baseline&nbsp;average</strong>.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!AFUR!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7194d61b-2e1b-4a73-8ad6-90b4b8823959_1024x647.png 424w, https://substackcdn.com/image/fetch/$s_!AFUR!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7194d61b-2e1b-4a73-8ad6-90b4b8823959_1024x647.png 848w, https://substackcdn.com/image/fetch/$s_!AFUR!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7194d61b-2e1b-4a73-8ad6-90b4b8823959_1024x647.png 1272w, https://substackcdn.com/image/fetch/$s_!AFUR!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7194d61b-2e1b-4a73-8ad6-90b4b8823959_1024x647.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!AFUR!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7194d61b-2e1b-4a73-8ad6-90b4b8823959_1024x647.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/7194d61b-2e1b-4a73-8ad6-90b4b8823959_1024x647.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!AFUR!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7194d61b-2e1b-4a73-8ad6-90b4b8823959_1024x647.png 424w, https://substackcdn.com/image/fetch/$s_!AFUR!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7194d61b-2e1b-4a73-8ad6-90b4b8823959_1024x647.png 848w, https://substackcdn.com/image/fetch/$s_!AFUR!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7194d61b-2e1b-4a73-8ad6-90b4b8823959_1024x647.png 1272w, https://substackcdn.com/image/fetch/$s_!AFUR!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7194d61b-2e1b-4a73-8ad6-90b4b8823959_1024x647.png 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p>The gains curve chart below illustrates that by focusing on the top ~10% of YC companies per our composite score, we capture ~4x as many Series A+ winners than we&#8217;d expect due to chance. While we&#8217;d capture even more winners in absolute terms if we invested beyond the top ~10% of each batch, we&#8217;d get diminishing returns in terms of Series A+ hit&nbsp;rate.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!wKmv!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffcaf1d2f-b015-4a2c-90de-a8b38a1b4c48_1024x690.png 424w, https://substackcdn.com/image/fetch/$s_!wKmv!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffcaf1d2f-b015-4a2c-90de-a8b38a1b4c48_1024x690.png 848w, https://substackcdn.com/image/fetch/$s_!wKmv!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffcaf1d2f-b015-4a2c-90de-a8b38a1b4c48_1024x690.png 1272w, https://substackcdn.com/image/fetch/$s_!wKmv!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffcaf1d2f-b015-4a2c-90de-a8b38a1b4c48_1024x690.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!wKmv!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffcaf1d2f-b015-4a2c-90de-a8b38a1b4c48_1024x690.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/fcaf1d2f-b015-4a2c-90de-a8b38a1b4c48_1024x690.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!wKmv!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffcaf1d2f-b015-4a2c-90de-a8b38a1b4c48_1024x690.png 424w, https://substackcdn.com/image/fetch/$s_!wKmv!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffcaf1d2f-b015-4a2c-90de-a8b38a1b4c48_1024x690.png 848w, https://substackcdn.com/image/fetch/$s_!wKmv!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffcaf1d2f-b015-4a2c-90de-a8b38a1b4c48_1024x690.png 1272w, https://substackcdn.com/image/fetch/$s_!wKmv!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffcaf1d2f-b015-4a2c-90de-a8b38a1b4c48_1024x690.png 1456w" sizes="100vw"></picture><div></div></div></div></figure></div><h3>The signals that matter&nbsp;most</h3><h3>1. Progress</h3><p>The single strongest predictor of future success we found is what our internal scoring system calls &#8220;Progress&#8221;&#8202;&#8212;&#8202;essentially, a structured AI-derived measure of how much real progress a company has made on its product and business by the time we evaluate it. Progress encompasses things like whether they&#8217;ve shipped a real product, whether they&#8217;re actively iterating with customers, whether they&#8217;re moving fast, and whether their progress feels proportionate to how long they&#8217;ve been at&nbsp;it.</p><p>Companies in the top 10% on Progress are several times more likely to reach Series A than a randomly chosen YC startup. It&#8217;s the single strongest signal in the entire&nbsp;dataset.</p><p>The intuition is straightforward: Series A investors aren&#8217;t betting on ideas, they&#8217;re betting on momentum. A company that has already shipped, learned, iterated, and shipped again is dramatically more likely to keep doing that, which is exactly what gets you to a Series A&nbsp;round.</p><h3>2. Product</h3><p>Closely behind Progress sits what we call Product&#8202;&#8212;&#8202;an AI-derived measure of how thoughtfully the actual product itself is built. Is the founder&#8217;s core insight a real one? Is the product elegant rather than bloated? Is there evidence of taste in how decisions were&nbsp;made?</p><p>Top-10% Product scores also deliver a lift several times above baseline.</p><p>It&#8217;s worth pausing on this for a moment. The two strongest qualitative signals we found are both about <em>the thing the founders have built</em>. It&#8217;s not about the market size, the cap table, or deal heat. It&#8217;s about the work itself and the founders&#8217; rate of progress on&nbsp;it.</p><h3>3. Revenue</h3><p>After Progress and Product, the next strongest signal is Revenue. This isn&#8217;t surprising since Revenue growth is downstream of Product and Progress, but the magnitude of the effect at even quite modest revenue tiers&nbsp;is.</p><p>Even a company with as little as $100K in ARR is more than twice as likely to reach Series A as a random YC startup. By $1M+ in ARR, you&#8217;re looking at a lift comparable to the strongest qualitative signals.</p><p>The takeaway for investors: don&#8217;t dismiss small-revenue YC startups as &#8220;too early&#8221; to evaluate. Even a few hundred thousand in ARR at the seed stage is a meaningfully positive signal. And for founders: getting to even modest paid traction before Demo Day puts you in a substantially different statistical bucket than companies that haven&#8217;t&nbsp;yet.</p><h3>4. Founder fundraising track&nbsp;record</h3><p>Founders who&#8217;ve previously raised significant capital for prior ventures show up as a real signal&#8202;&#8212;&#8202;though more modest than the ones above. Specifically, we found that the more capital a founding team has previously raised, the more likely their current YC startup is to reach Series&nbsp;A.</p><p>A subtle but important point: this is &#8220;have you previously <em>raised</em> capital,&#8221; not &#8220;have you previously <em>exited</em>.&#8221; The two are correlated but distinct, and prior fundraising turns out to be the stronger of the two signals at the YC Demo Day&nbsp;stage.</p><h3>5. Rebel Theorem&nbsp;scores</h3><p>Unsurprisingly, our Rebel Theorem scores also produce a significant lift for companies scoring well. It&#8217;s a useful signal, especially given that it operates on every company objectively based on hundreds of discrete data points, and these scores are <em>additive</em> to our newer AI-derived scores.</p><p>But the more interesting finding is that the Rebel Theorem score is asymmetric. The lift at the top end is moderate, but the <em>downside signal</em> at the bottom end is sharper: companies with very low scores reach Series A at rates well below the baseline. In other words, Rebel Theorem is a stronger filter for ruling weak companies <em>out</em> than for pulling strong companies <em>in</em>.</p><p>That&#8217;s an important insight for any data-driven investor building this kind of model. A score doesn&#8217;t have to be a perfect oracle of upside to be valuable&#8202;&#8212;&#8202;being a reliable filter on the downside is, on its own, a hugely useful property at the seed stage, where avoiding losers is half the&nbsp;battle.</p><h3>6. Team&nbsp;quality</h3><p>Now to the result that surprised us most. Of the dozen or so AI scoring dimensions we assess, the one called Team&#8202;&#8212;&#8202;covering things like founder pedigree, technical skill, prior experience, and complementary co-founder skill sets&#8202;&#8212;&#8202;produced only a middling&nbsp;lift.</p><p>When we dove in deeper, we found the situation is actually quite nuanced. It&#8217;s not that Team scores don&#8217;t matter, it&#8217;s that they only matter at the extreme. In other words, <em>really</em> good teams perform extraordinarily well, but nearly ok or good teams don&#8217;t do much better than&nbsp;average.</p><p>Bear in mind that YC&#8217;s own admissions process is, among other things, an exceptionally rigorous filter on team quality. YC&#8217;s admissions rate is &lt;1%, so by the time a company shows up in a YC batch, the variance in team quality across the cohort is already compressed dramatically relative to the broader founder population. The Team signal still discriminates within that compressed range&#8202;&#8212;&#8202;just less powerfully than dimensions like Progress or Product, where there&#8217;s more variance to work&nbsp;with.</p><p>This squares with something we&#8217;ve believed for a long time at Rebel: at the YC Demo Day stage, the question isn&#8217;t really &#8220;is this a good team?&#8221; since most of them are. The question is &#8220;is this an <em>exceptional</em> team and are they executing quickly on the right&nbsp;thing?&#8221;</p><h3>The signals that don&#8217;t matter (as&nbsp;much)</h3><p>Just as interesting as the signals that work are the ones that don&#8217;t. A few worth calling&nbsp;out:</p><h3>Launch timing</h3><p>How early a company launches publicly before Demo Day has little predictive value, with a notable exception when companies launch <em>very</em> early in the batch. This probably reflects the fact that companies who get organized early are more likely to be high-Progress companies&#8202;&#8212;&#8202;which we&#8217;re already capturing directly. Once you control for the other signals, launch timing doesn&#8217;t tell you much, but the statistical uptick in the performance of companies that launch early is an interesting phenomenon.</p><h3>Demo Day&nbsp;&#8220;heat&#8221;</h3><p>I covered this in detail in <a href="https://medium.com/p/on-investing-in-hot-y-combinator-startups-aed26869f77c">On investing in &#8220;hot&#8221; Y Combinator startups</a>, so I&#8217;ll keep it brief here. Demo Day investor enthusiasm has a real but quite weak correlation with eventual outcomes. Given the higher entry valuations that hot deals command, the effect is roughly washed out by the time you account for price. In other words: deal heat is at best a weak positive signal that gets fully priced in by the market. It&#8217;s not a useful basis for an allocation decision.</p><h3>The compounding effect</h3><p>One last observation, which is arguably the most important methodological finding of all: <strong>the signals compound</strong>. Companies where two or more of the strong signals fire together are dramatically more likely to succeed than companies where any one fires&nbsp;alone.</p><p>We now screen companies using a combinatorial ML/AI scoring system with hundreds of variables precisely because of this. The marginal value of any single signal is real but limited&#8202;&#8212;&#8202;the value of precisely weighted combinations of signals is <em>much</em>&nbsp;greater.</p><p>This is one of the reasons that, despite all the recent hype around AI in venture, only ~1% of funds today employ truly rigorous and backtested data-driven investment processes. It&#8217;s hard work, and it&#8217;s only valuable if you have the right data and team to train a sophisticated multi-variate model, which took us years and millions of R&amp;D&nbsp;dollars.</p><p>Knowing which signals drive YC startup performance is not sufficient on its own. You also need the technical infrastructure to collect vast amounts of data on every YC startup and founder, know how much to weight each variable, and then score and rank them in an objective and scalable&nbsp;way.</p><h3>What this&nbsp;means</h3><p>For investors at the YC Demo Day stage, the implications are pretty clear. If you&#8217;re not weighting product progress and revenue traction at least as heavily as team pedigree, you&#8217;re probably mispricing the cohort. The instinct to &#8220;bet on the team&#8221; is right at the population level, but at the YC stage, where team variance is already compressed by selection, it&#8217;s the product itself and rate of progress that does the real discriminating.</p><p>For founders, the message is one I&#8217;ve made before but worth restating: the most valuable thing you can do in the weeks before Demo Day is ship visible product progress and get to real paying customers. Both move the statistical needle far more than the things founders typically obsess over&#8202;&#8212;&#8202;pitch polish, deck design, networking, who&#8217;s &#8220;hot&#8221; this batch,&nbsp;etc.</p><p>For us at Rebel, the underlying lesson is that even an extremely sophisticated, data-driven approach to YC startup investing must continually evolve. Every six months or so, we retest our priors against the latest cohort of outcomes data. Some signals get stronger, some weaker, and some may fall out entirely. The whole point of building the world&#8217;s most comprehensive YC dataset and ML/AI scoring system trained on it, is that we can keep doing exactly that, indefinitely.</p><p>&#185; &#8220;Series A or later&#8221; is the success threshold I use throughout this post, both because it&#8217;s a clean and observable milestone that we can assess in relatively recent batches. What ultimately drives returns for a seed-stage investors like us are really the unicorns, but it&#8217;s still a useful intermediate filter. While only a minority of the Series A startups ultimately become unicorns, every unicorn raised a Series A at some&nbsp;point.</p>]]></content:encoded></item><item><title><![CDATA[On the freakishly strong YC W26 batch]]></title><description><![CDATA[Now that Rebel Fund has invested in nearly 300 top Y Combinator startups, I&#8217;m often asked by our LPs and other investors what I think about the latest YC batch.]]></description><link>https://insights.rebelfund.vc/p/on-the-freakishly-strong-yc-w26-batch-056ccb666076</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-the-freakishly-strong-yc-w26-batch-056ccb666076</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Mon, 09 Mar 2026 23:55:36 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/3bc7e9f8-21f2-4398-bce3-37114a81aaac_1024x434.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>Now that <a href="https://www.rebelfund.vc/">Rebel Fund</a> has invested in nearly 300 top <a href="https://www.ycombinator.com/">Y Combinator</a> startups, I&#8217;m often asked by our LPs and other investors what I think about the latest YC batch. Since I&#8217;ve been to every YC Demo Day since my own in 2013, I could answer this question subjectively, but I prefer to answer it with data &#8212;and the data shows the YC Winter 2026 (W26) batch is freakishly strong.</p><p>As my long-time readers know, Rebel has invested millions of dollars into developing the world&#8217;s most sophisticated ML/AI algorithm for predicting YC startup success, <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">Rebel Theorem 4.0</a>, which we use not only to accurately assess the quality of individual YC startups, but also <em>entire YC&nbsp;batches</em>.</p><p>To illustrate the strength of W26, this chart shows the percentage of startups in each YC batch over the past few years that are &#8220;top 20% companies&#8221; as scored by our algorithm (benchmarked to pre-W26 data). A whopping 35% of W26 startups meet this threshold, far more than any other YC&nbsp;batch.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Bss_!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd41b0532-6c8a-41bf-93be-9cb91592d1a1_1024x434.jpeg 424w, https://substackcdn.com/image/fetch/$s_!Bss_!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd41b0532-6c8a-41bf-93be-9cb91592d1a1_1024x434.jpeg 848w, https://substackcdn.com/image/fetch/$s_!Bss_!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd41b0532-6c8a-41bf-93be-9cb91592d1a1_1024x434.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!Bss_!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd41b0532-6c8a-41bf-93be-9cb91592d1a1_1024x434.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Bss_!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd41b0532-6c8a-41bf-93be-9cb91592d1a1_1024x434.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/d41b0532-6c8a-41bf-93be-9cb91592d1a1_1024x434.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!Bss_!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd41b0532-6c8a-41bf-93be-9cb91592d1a1_1024x434.jpeg 424w, https://substackcdn.com/image/fetch/$s_!Bss_!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd41b0532-6c8a-41bf-93be-9cb91592d1a1_1024x434.jpeg 848w, https://substackcdn.com/image/fetch/$s_!Bss_!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd41b0532-6c8a-41bf-93be-9cb91592d1a1_1024x434.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!Bss_!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd41b0532-6c8a-41bf-93be-9cb91592d1a1_1024x434.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p>In case you&#8217;re wondering if there&#8217;s just a small group of top startups holding up W26 batch quality, that&#8217;s apparently not the case. As you&#8217;ll see in the charts below, the <em>entire distribution curve</em> of the batch has shifted in the right direction, with significantly more startups predicted to be successful, and significantly less predicted to die or become zombies, simultaneously at the 25th, 50th, and 75th percentiles.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!hs13!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbaa9ecce-324e-48b5-be50-e29951af98c7_1024x320.jpeg 424w, https://substackcdn.com/image/fetch/$s_!hs13!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbaa9ecce-324e-48b5-be50-e29951af98c7_1024x320.jpeg 848w, https://substackcdn.com/image/fetch/$s_!hs13!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbaa9ecce-324e-48b5-be50-e29951af98c7_1024x320.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!hs13!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbaa9ecce-324e-48b5-be50-e29951af98c7_1024x320.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!hs13!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbaa9ecce-324e-48b5-be50-e29951af98c7_1024x320.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/baa9ecce-324e-48b5-be50-e29951af98c7_1024x320.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!hs13!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbaa9ecce-324e-48b5-be50-e29951af98c7_1024x320.jpeg 424w, https://substackcdn.com/image/fetch/$s_!hs13!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbaa9ecce-324e-48b5-be50-e29951af98c7_1024x320.jpeg 848w, https://substackcdn.com/image/fetch/$s_!hs13!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbaa9ecce-324e-48b5-be50-e29951af98c7_1024x320.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!hs13!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbaa9ecce-324e-48b5-be50-e29951af98c7_1024x320.jpeg 1456w" sizes="100vw"></picture><div></div></div></div></figure></div><p>This begs the question &#8220;What makes the W26 batch so strong?&#8221; Of course my crack Rebel data science team has some&nbsp;answers.</p><p>The main thing is that W26 batch founders demonstrate more of the characteristics that have historically predicted strong YC startup outcomes. This is probably no coincidence, since as I mentioned in my previous post <a href="https://jaredheyman.medium.com/on-the-new-y-combinator-3c28e548896c">On the &#8216;new&#8217; Combinator</a>, YC under Garry Tan&#8217;s leadership has generally refocused on all the things that made it great in the first&nbsp;place.</p><p>For example, we&#8217;ve seen in our data that younger founders are generally more likely to be successful, and W26 founders are much younger and fresher out of school than their predecessors.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!AjNQ!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15d26257-9293-4e88-9547-918219e9b8dc_1024x390.jpeg 424w, https://substackcdn.com/image/fetch/$s_!AjNQ!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15d26257-9293-4e88-9547-918219e9b8dc_1024x390.jpeg 848w, https://substackcdn.com/image/fetch/$s_!AjNQ!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15d26257-9293-4e88-9547-918219e9b8dc_1024x390.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!AjNQ!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15d26257-9293-4e88-9547-918219e9b8dc_1024x390.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!AjNQ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15d26257-9293-4e88-9547-918219e9b8dc_1024x390.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/15d26257-9293-4e88-9547-918219e9b8dc_1024x390.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!AjNQ!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15d26257-9293-4e88-9547-918219e9b8dc_1024x390.jpeg 424w, https://substackcdn.com/image/fetch/$s_!AjNQ!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15d26257-9293-4e88-9547-918219e9b8dc_1024x390.jpeg 848w, https://substackcdn.com/image/fetch/$s_!AjNQ!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15d26257-9293-4e88-9547-918219e9b8dc_1024x390.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!AjNQ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15d26257-9293-4e88-9547-918219e9b8dc_1024x390.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>As another example, we&#8217;ve seen that Bay Area companies are more likely to be successful than, say, European companies, and sure enough, W26 companies are much more concentrated in the Bay&nbsp;Area.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!quza!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F823c0853-2b60-4a9f-b4bf-948ca83ee87c_1024x385.jpeg 424w, https://substackcdn.com/image/fetch/$s_!quza!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F823c0853-2b60-4a9f-b4bf-948ca83ee87c_1024x385.jpeg 848w, https://substackcdn.com/image/fetch/$s_!quza!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F823c0853-2b60-4a9f-b4bf-948ca83ee87c_1024x385.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!quza!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F823c0853-2b60-4a9f-b4bf-948ca83ee87c_1024x385.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!quza!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F823c0853-2b60-4a9f-b4bf-948ca83ee87c_1024x385.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/823c0853-2b60-4a9f-b4bf-948ca83ee87c_1024x385.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!quza!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F823c0853-2b60-4a9f-b4bf-948ca83ee87c_1024x385.jpeg 424w, https://substackcdn.com/image/fetch/$s_!quza!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F823c0853-2b60-4a9f-b4bf-948ca83ee87c_1024x385.jpeg 848w, https://substackcdn.com/image/fetch/$s_!quza!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F823c0853-2b60-4a9f-b4bf-948ca83ee87c_1024x385.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!quza!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F823c0853-2b60-4a9f-b4bf-948ca83ee87c_1024x385.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Looking at industry-related features, we&#8217;ve seen that consumer companies are less likely to be successful than, say, industrials companies, and W26 has much less consumer companies and more industrials companies.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!-dpZ!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb0cb3e31-c4c7-4429-824f-59e692d6f4f1_1024x390.jpeg 424w, https://substackcdn.com/image/fetch/$s_!-dpZ!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb0cb3e31-c4c7-4429-824f-59e692d6f4f1_1024x390.jpeg 848w, https://substackcdn.com/image/fetch/$s_!-dpZ!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb0cb3e31-c4c7-4429-824f-59e692d6f4f1_1024x390.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!-dpZ!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb0cb3e31-c4c7-4429-824f-59e692d6f4f1_1024x390.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!-dpZ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb0cb3e31-c4c7-4429-824f-59e692d6f4f1_1024x390.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/b0cb3e31-c4c7-4429-824f-59e692d6f4f1_1024x390.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!-dpZ!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb0cb3e31-c4c7-4429-824f-59e692d6f4f1_1024x390.jpeg 424w, https://substackcdn.com/image/fetch/$s_!-dpZ!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb0cb3e31-c4c7-4429-824f-59e692d6f4f1_1024x390.jpeg 848w, https://substackcdn.com/image/fetch/$s_!-dpZ!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb0cb3e31-c4c7-4429-824f-59e692d6f4f1_1024x390.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!-dpZ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb0cb3e31-c4c7-4429-824f-59e692d6f4f1_1024x390.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>To be clear, I&#8217;m greatly over-simplifying the examples above, since these are just a few of the hundreds of features that <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">Rebel Theorem 4.0</a> assesses in parallel for each startup, and the interaction <em>between</em> features matters much more than any feature alone. In other words, while younger founders generally have higher odds of success, that&#8217;s not true in every scenario. The same goes for Bay Area companies, industrials companies, etc.</p><p>There are many more ways that W26 founders and companies more closely reflect the historical ideal of what a top YC startup should look like, from founder pedigree, to prior founder accomplishments, to <a href="https://jaredheyman.medium.com/on-founder-product-fit-7f1ad68d7af5">founder-product fit</a>, and some ways they don&#8217;t. This post just focuses on the macro&nbsp;trend.</p><p>The billion dollar question on my mind and presumably the YC partners&#8217; is whether in today&#8217;s world of Claude Code, Codex, etc if the fundamentals of building and managing a startup have changed so dramatically that the technical pedigree, educational and work background, skills, personality traits, etc that have historically predicted YC founder success still apply. I don&#8217;t have a confident answer to that yet, but since <a href="https://www.rebelfund.vc/">Rebel Fund</a> tracks hundreds of characteristics and outcomes of every YC company and founder in nearly real-time, once the data is in, we&#8217;ll be the first to&nbsp;know.</p>]]></content:encoded></item><item><title><![CDATA[On hitting the “bullseye”]]></title><description><![CDATA[It seems the most controversial topic in early-stage venture today is whether a large diversified portfolio strategy (70+ companies) or a small concentrated portfolio strategy (10&#8211;20 companies) is better for LPs.]]></description><link>https://insights.rebelfund.vc/p/on-hitting-the-bullseye-6295e99d2a89</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-hitting-the-bullseye-6295e99d2a89</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Fri, 07 Nov 2025 19:46:12 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/d59c909e-23d1-4a65-b62a-edc52a223e72_799x803.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>It seems the most controversial topic in early-stage venture today is whether a large diversified portfolio strategy (70+ companies) or a small concentrated portfolio strategy (10&#8211;20 companies) is better for&nbsp;LPs.</p><p>At <a href="https://www.rebelfund.vc/">Rebel Fund</a>, we confidently chose the former, targeting 150+ <a href="https://www.ycombinator.com/\">Y Combinator</a> startups per fund, or the top 10% of YC startups driving ~96% of all investor returns. The math is clear that a large portfolio strategy is superior when investing in YC startups, especially after accounting for risk. However, we still run across many institutional LPs who believe otherwise.</p><p>In my previous post on this topic <a href="https://link.medium.com/D6Ootw8LdJb">On the power law of Y Combinator startups</a> I share the power law curve that drives YC startup returns, and the results of several Monte Carlo simulations we ran proving that a large portfolio strategy is optimal when it comes to YC startup investing. In this post, I&#8217;ll take it a step further with the help of a visual aid that anyone who been to a gun range or Irish pub is familiar with: <em><strong>a&nbsp;bullseye</strong></em></p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Nhu2!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5dccf63-9e15-440e-b9d1-8fc17692030c_799x803.jpeg 424w, https://substackcdn.com/image/fetch/$s_!Nhu2!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5dccf63-9e15-440e-b9d1-8fc17692030c_799x803.jpeg 848w, https://substackcdn.com/image/fetch/$s_!Nhu2!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5dccf63-9e15-440e-b9d1-8fc17692030c_799x803.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!Nhu2!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5dccf63-9e15-440e-b9d1-8fc17692030c_799x803.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Nhu2!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5dccf63-9e15-440e-b9d1-8fc17692030c_799x803.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/d5dccf63-9e15-440e-b9d1-8fc17692030c_799x803.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!Nhu2!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5dccf63-9e15-440e-b9d1-8fc17692030c_799x803.jpeg 424w, https://substackcdn.com/image/fetch/$s_!Nhu2!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5dccf63-9e15-440e-b9d1-8fc17692030c_799x803.jpeg 848w, https://substackcdn.com/image/fetch/$s_!Nhu2!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5dccf63-9e15-440e-b9d1-8fc17692030c_799x803.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!Nhu2!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5dccf63-9e15-440e-b9d1-8fc17692030c_799x803.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div><figcaption class="image-caption">All YC Startups (2005&#8211;2018)</figcaption></figure></div><p>In this bullseye, each circle represents a different startup valuation range, drawn to scale to illustrate how common each one is amongst Y Combinator startups at least 7+ years old. <a href="https://www.rebelfund.vc/">Rebel Fund</a> maintains the largest database that exists of YC startups, founders &amp; outcomes outside of YC itself, so we&#8217;re in a unique position to share these statistics:</p><p><strong>Dark red &#8220;decacorn&#8221; circle </strong>&#8594; $10B+ valuation (~1% of YC startups, median valuation ~$20B)</p><p><strong>Red &#8220;unicorn&#8221; circle </strong>&#8594; $1B-$9.9B valuation (~6% of YC startups, median valuation ~$2B)</p><p><strong>Pink &#8220;$100M+&#8221; circle</strong> &#8594; $100M-$999M valuation (~20% of YC startups, median valuation ~$400M)</p><p><strong>Grey &#8220;non-outlier&#8221; circle</strong> &#8594; &lt;$100M valuation (~80% of YC startups, median valuation ~$10M)</p><p>As a GP, my job is to fire shots at this bullseye that land as close to the center as possible. The bad news for me is that hitting the red circles is very difficult, but the good news is the payoff is huge when I hit&nbsp;them.</p><p>Below are estimated net return multiples for a YC startup investor hitting each circle, assuming a median ~$20M entry valuation and ~50% dilution to&nbsp;exit:</p><p><strong>Dark red &#8220;decacorn&#8221; circle </strong>&#8594;&nbsp;~500x</p><p><strong>Red &#8220;unicorn&#8221; circle &#8594;&nbsp;~</strong>50x</p><p><strong>Pink &#8220;$100M+&#8221; circle &#8594;&nbsp;~</strong>10x</p><p><strong>Grey &#8220;non-outlier&#8221; circle</strong> &#8212;&gt; ~0x (to be conservative)</p><p>I based these return multiples on median rather than average outcomes to be conservative, as averages are much higher due to the mega-outliers.</p><p>To help illustrate what different portfolio strategies (i.e., number of shots fired) mean for LPs, I built <a href="https://colab.research.google.com/drive/1yTwKZqoPDru6G85FzPKwJrG339EWdoVs?usp=sharing">a Python script</a> that generates customized bullseyes based on various inputs around the investment universe, and randomly fires shots at it based on certain portfolio size and investment accuracy assumptions. Each bullseye essentially shows a single trial of a Monte Carlo simulation as featured in my <a href="https://link.medium.com/D6Ootw8LdJb">power law post</a>, but I think this visual will make it easier for non-statisticians to understand.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!LKrd!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fed2d005d-b39f-47ae-bebc-ca0c5787a974_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!LKrd!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fed2d005d-b39f-47ae-bebc-ca0c5787a974_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!LKrd!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fed2d005d-b39f-47ae-bebc-ca0c5787a974_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!LKrd!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fed2d005d-b39f-47ae-bebc-ca0c5787a974_1024x1024.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!LKrd!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fed2d005d-b39f-47ae-bebc-ca0c5787a974_1024x1024.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/ed2d005d-b39f-47ae-bebc-ca0c5787a974_1024x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!LKrd!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fed2d005d-b39f-47ae-bebc-ca0c5787a974_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!LKrd!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fed2d005d-b39f-47ae-bebc-ca0c5787a974_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!LKrd!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fed2d005d-b39f-47ae-bebc-ca0c5787a974_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!LKrd!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fed2d005d-b39f-47ae-bebc-ca0c5787a974_1024x1024.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">Diversified YC Seed&nbsp;Strategy</figcaption></figure></div><p>This first bullseye illustrates our portfolio strategy at Rebel. We make 150+ investments per fund, exclusively in YC startups, and I&#8217;ve assumed conservatively our accuracy is 10% on a scale from 0% (same as chance) &#8594; 100% (perfection).</p><p>In this simulation, even though ~70% of our shots don&#8217;t hit the mark, we still hit 4 decacorns, 10 unicorns, and twenty-nine $100M+ companies, resulting in an ~18x overall net return multiple&#8202;&#8212;&#8202;not&nbsp;bad!</p><p>Now, let&#8217;s compare that with more concentrated, but still YC-focused investment strategy:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!MLNq!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa4cf7375-243d-4016-b0c2-0c39dfcbd6a1_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!MLNq!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa4cf7375-243d-4016-b0c2-0c39dfcbd6a1_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!MLNq!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa4cf7375-243d-4016-b0c2-0c39dfcbd6a1_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!MLNq!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa4cf7375-243d-4016-b0c2-0c39dfcbd6a1_1024x1024.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!MLNq!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa4cf7375-243d-4016-b0c2-0c39dfcbd6a1_1024x1024.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/a4cf7375-243d-4016-b0c2-0c39dfcbd6a1_1024x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!MLNq!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa4cf7375-243d-4016-b0c2-0c39dfcbd6a1_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!MLNq!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa4cf7375-243d-4016-b0c2-0c39dfcbd6a1_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!MLNq!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa4cf7375-243d-4016-b0c2-0c39dfcbd6a1_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!MLNq!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa4cf7375-243d-4016-b0c2-0c39dfcbd6a1_1024x1024.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">Concentrated YC Seed&nbsp;Strategy</figcaption></figure></div><p>This bullseye includes only 10 investments in YC startups, with the same 10% accuracy assumption as&nbsp;above.</p><p>In this simulation, again ~70% of the shots landed in the grey, but there were 2 unicorn investments and one $100M+ company investment, resulting in an 11x net return multiple&#8202;&#8212;&#8202;still quite&nbsp;good!</p><p>What&#8217;s interesting about this simulation is the unicorn rate was very lucky at 20%, over 3x the YC average, yet it <em>still</em> underperformed Rebel&#8217;s large portfolio strategy. The reason is that no decacorns were hit, which is unsurprising since they&#8217;re just ~1% of YC startups&#8202;&#8212;&#8202;a hard mark to hit with just 10 shots. Even worse, if just one of those unicorn hits landed in the grey, the net return multiple would be cut in half... it&#8217;s a fragile portfolio.</p><p>Next let&#8217;s see what happens if the investment universe is different.</p><p>Outside of YC, the venture-backed startup unicorn rate is 1%-2% according to our research. Let&#8217;s peg it at 1.5% and generously assume the ratio of decacorns to unicorns to $100M+ companies is the same as in YC. We&#8217;ll also generously assume net return multiples for each outcome category are twice as high since entry valuations are about half of YC companies&#8217;.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!2lC2!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff7720df8-bcae-4426-9ab7-1abfbc03de74_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!2lC2!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff7720df8-bcae-4426-9ab7-1abfbc03de74_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!2lC2!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff7720df8-bcae-4426-9ab7-1abfbc03de74_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!2lC2!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff7720df8-bcae-4426-9ab7-1abfbc03de74_1024x1024.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!2lC2!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff7720df8-bcae-4426-9ab7-1abfbc03de74_1024x1024.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/f7720df8-bcae-4426-9ab7-1abfbc03de74_1024x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!2lC2!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff7720df8-bcae-4426-9ab7-1abfbc03de74_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!2lC2!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff7720df8-bcae-4426-9ab7-1abfbc03de74_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!2lC2!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff7720df8-bcae-4426-9ab7-1abfbc03de74_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!2lC2!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff7720df8-bcae-4426-9ab7-1abfbc03de74_1024x1024.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">Diversified Non-YC Seed&nbsp;Strategy</figcaption></figure></div><p>Now that our investment universe is non-YC, the red circles are much smaller, so even with the same 150 shots and 10% accuracy, the results are <em>much</em> worse, even after accounting for the lower entry valuations.</p><p>In this simulation there were no decacorn hits, 3 unicorn hits, and 10 hits of $100M+ companies&#8202;&#8212;&#8202;all in line with what we&#8217;d expect to see statistically. The net return multiple is ~3x, which is fairly typical for a generalist diversified seed&nbsp;fund.</p><p>This bullseye highlights why a portfolio strategy that works very well in one investment universe, elite YC startups, may not work well in&nbsp;another.</p><p>Now let&#8217;s see what happens if our non-YC seed investor only takes 10&nbsp;shots.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!fc9I!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20ac18fa-66e7-4650-b638-c2271e69d829_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!fc9I!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20ac18fa-66e7-4650-b638-c2271e69d829_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!fc9I!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20ac18fa-66e7-4650-b638-c2271e69d829_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!fc9I!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20ac18fa-66e7-4650-b638-c2271e69d829_1024x1024.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!fc9I!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20ac18fa-66e7-4650-b638-c2271e69d829_1024x1024.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/20ac18fa-66e7-4650-b638-c2271e69d829_1024x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!fc9I!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20ac18fa-66e7-4650-b638-c2271e69d829_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!fc9I!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20ac18fa-66e7-4650-b638-c2271e69d829_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!fc9I!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20ac18fa-66e7-4650-b638-c2271e69d829_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!fc9I!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20ac18fa-66e7-4650-b638-c2271e69d829_1024x1024.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">Concentrated Non-YC Seed&nbsp;Strategy</figcaption></figure></div><p>The only difference between this simulation and the previous one is the number of shots fired, but the results look dramatically different.</p><p>In this simulation, there were no decacorns or unicorns hit despite the same investor accuracy assumption as before, simply because it&#8217;s unlikely to hit such a small target with just 10 shots. This investor hit a single $100M+ company to help generate a 2x overall net return multiple, which is enough to return LP capital after fees, but nothing to write home&nbsp;about.</p><p>Ironically, it&#8217;s perfectly rational for an emerging manager to assume the higher risk of a concentrated portfolio strategy even <em>knowing</em> that its expected returns are worse than with a diversified portfolio strategy. Why?</p><p>Because if he somehow beats the odds and hits a couple of unicorns (or especially a decacorn) with just 10&#8211;20 investments, then his performance is spectacular and he can easily raise another couple funds off that early success. If he <em>doesn&#8217;t</em> beat the odds and his returns are mediocre, then he still walks away with a couple million in management fees, plus it&#8217;s much easier to make 10&#8211;20 investments than 100+ per fund. Don&#8217;t shoot the messenger &#128514;</p><p>In all the bullseyes shared thus far, I kept investor accuracy constant at 10%. However in the real world this is an important variable, with a major impact on the ideal portfolio strategy for a given investor.</p><p>Let&#8217;s imagine I have a god-like ability to see the future and my investor accuracy is 100%. Then I would make exactly one investment in each YC batch in the biggest decacorn and my bullseye would look like&nbsp;this:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!__LI!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5567c7d1-cf85-48b7-bad4-938787e5aec4_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!__LI!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5567c7d1-cf85-48b7-bad4-938787e5aec4_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!__LI!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5567c7d1-cf85-48b7-bad4-938787e5aec4_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!__LI!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5567c7d1-cf85-48b7-bad4-938787e5aec4_1024x1024.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!__LI!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5567c7d1-cf85-48b7-bad4-938787e5aec4_1024x1024.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/5567c7d1-cf85-48b7-bad4-938787e5aec4_1024x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!__LI!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5567c7d1-cf85-48b7-bad4-938787e5aec4_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!__LI!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5567c7d1-cf85-48b7-bad4-938787e5aec4_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!__LI!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5567c7d1-cf85-48b7-bad4-938787e5aec4_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!__LI!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5567c7d1-cf85-48b7-bad4-938787e5aec4_1024x1024.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">God&#8217;s YC Seed&nbsp;Strategy</figcaption></figure></div><p>In this happy scenario, all of my ultra-concentrated funds return over 500x and I ride off into the sunset in a private jet of solid gold. Of course this is a fantasy, and not just because gold is too soft and heavy for jet construction.</p><p>No mortal can predict which seed-stage startups will become unicorns, much less decacorns, with a high degree of accuracy. At Rebel, we have the world&#8217;s most sophisticated ML/AI algorithm for predicting YC startup success, <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">Rebel Theorem 4.0</a>, with hundreds of features trained on millions of YC data points, plus a team of <a href="https://www.rebelfund.vc/#section-partners">famous YC unicorn founders</a> as partners, and we <em>still</em> miss great YC startups all the&nbsp;time.</p><p>The best any investor can hope for at the seed stage is to outperform averages for their target investment universe, and most importantly, avoid adverse selection, which is the <em>real</em> killer of seed portfolios. You would think the worst an investor&#8217;s accuracy can be is 0 (i.e., same as chance) but that&#8217;s not true at all. It can easily go negative thanks to adverse selection.</p><p>Imagine you show up to YC Demo Day for the first time with a fat checkbook and high hopes for investing in tomorrow&#8217;s decacorns. Let&#8217;s also assume you&#8217;ve read this blog post and know the importance of portfolio diversification, so you&#8217;re ready to invest in 20 startups in just that YC&nbsp;batch.</p><p>Assuming zero skill (i.e., investor accuracy) but perfect deal access, we&#8217;d expect about about one of your investments to turn into unicorns, a couple to turn into $100M+ companies, and the rest to be non-outliers. In other words, your bullseye for that batch <em>should</em> look something like&nbsp;this:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!pWlC!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1af340e3-3033-4950-bed9-8c94e585dc85_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!pWlC!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1af340e3-3033-4950-bed9-8c94e585dc85_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!pWlC!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1af340e3-3033-4950-bed9-8c94e585dc85_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!pWlC!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1af340e3-3033-4950-bed9-8c94e585dc85_1024x1024.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!pWlC!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1af340e3-3033-4950-bed9-8c94e585dc85_1024x1024.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/1af340e3-3033-4950-bed9-8c94e585dc85_1024x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!pWlC!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1af340e3-3033-4950-bed9-8c94e585dc85_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!pWlC!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1af340e3-3033-4950-bed9-8c94e585dc85_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!pWlC!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1af340e3-3033-4950-bed9-8c94e585dc85_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!pWlC!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1af340e3-3033-4950-bed9-8c94e585dc85_1024x1024.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">YC Demo Day with Perfect Deal&nbsp;Access</figcaption></figure></div><p>However, since the top startups in each YC batch end up with heavily oversubscribed rounds, the founders get to choose their investors, and they&#8217;re probably not going to choose an unknown investor who can&#8217;t add significant value to their company beyond capital. So, your <em>effective</em> investor accuracy is negative, and your bullseye may actually look like&nbsp;this:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!F_6Q!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F93ba46b2-8265-4617-a039-d2118325ce68_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!F_6Q!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F93ba46b2-8265-4617-a039-d2118325ce68_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!F_6Q!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F93ba46b2-8265-4617-a039-d2118325ce68_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!F_6Q!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F93ba46b2-8265-4617-a039-d2118325ce68_1024x1024.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!F_6Q!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F93ba46b2-8265-4617-a039-d2118325ce68_1024x1024.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/93ba46b2-8265-4617-a039-d2118325ce68_1024x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!F_6Q!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F93ba46b2-8265-4617-a039-d2118325ce68_1024x1024.png 424w, https://substackcdn.com/image/fetch/$s_!F_6Q!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F93ba46b2-8265-4617-a039-d2118325ce68_1024x1024.png 848w, https://substackcdn.com/image/fetch/$s_!F_6Q!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F93ba46b2-8265-4617-a039-d2118325ce68_1024x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!F_6Q!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F93ba46b2-8265-4617-a039-d2118325ce68_1024x1024.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div><figcaption class="image-caption">YC Demo Day with Adverse Selection</figcaption></figure></div><p>In this unfortunate but common scenario, your investor return is zero, despite having done nothing wrong in terms of selection. In fact, adverse selection is so dangerous that even the best shooting skills in the world can&#8217;t overcome&nbsp;it.</p><p>That&#8217;s why I built <a href="https://www.rebelfund.vc/">Rebel Fund</a> from the ground up to optimize for deal access. Every choice I made in the early days, from partner selection to portfolio support to target check size to investment timing, was made to assure we get into every YC company we want. It&#8217;s paid off&#8202;&#8212;&#8202;after hundreds of investments our deal win rate is nearly&nbsp;100%.</p><p>I could generate more bullseyes to show the pros and cons of different portfolio strategies across different investment universes and even stages (Series A, Series B, etc) but since I&#8217;ve open-sourced <a href="https://colab.research.google.com/drive/1yTwKZqoPDru6G85FzPKwJrG339EWdoVs?usp=sharing">my Python script</a>, you can do so yourself.</p><p>Now let me address some questions that anticipate hearing from institutional LPs who read this&nbsp;post:</p><blockquote><p>What about investor ownership targets? Doesn&#8217;t that matter&nbsp;too?</p></blockquote><p>Nope. There is no difference in any fund performance metric (IRR, TVPI, MOIC, DPI, etc) between a smaller fund that owns 2% or a larger fund that owns 20% of the same underlying portfolio companies on the same deal terms. At the end of the day, all that matters is the number of investments (i.e., risk) and average net return multiple (i.e., return). My cynical take is fund managers only have minimum ownership targets to justify a larger fund size &#8594; more management fees.</p><blockquote><p>We invest across 20 venture funds. If each of them only invests in 10&#8211;20 companies, don&#8217;t we get enough diversification across the 200&#8211;400 combined underlying portfolio companies?</p></blockquote><p>Your venture portfolio performance is the weighted average of all your individual fund investments, so you every fund you invest in should have the optimal portfolio strategy for their investment universe, stage, and GP skill/access level (i.e., investor accuracy). A bunch of sub-optimal portfolio strategies don&#8217;t magically become optimal once combined.</p><p>Yes, if you invest in a bunch of a concentrated portfolio funds then it&#8217;s possible that a few lucky GPs could carry the rest, but why take the risk? It&#8217;s safer to simply optimize the quality of every fund investment you make, and a big part of that is &#8220;portfolio strategy fit&#8221; for their unique&nbsp;thesis.</p><blockquote><p>If we invest across 20 venture funds with diversified portfolio strategies, we&#8217;d have thousands of underlying portfolio companies. How can I can I outperform the venture market and keep my job with so much diversification?</p></blockquote><p>There are over 3,000 venture funds in the US alone, and you want to invest in what&#8230; 20 of them? You couldn&#8217;t &#8220;index&#8221; the venture market even if you tried. Besides, every venture fund has its own unique thesis, investment universe, manager skill &amp; biases, etc. all bringing you unique exposure.</p><blockquote><p>But [insert famous fund] had a concentrated portfolio strategy and returned [insert amazing multiple] to LPs! How could that be if a diversified portfolio strategy is superior?</p></blockquote><p>Of course, if a manager only makes 10&#8211;20 investments and a few of them land in the center rings of the bullseye, their numbers are fantastic. Just remember how small those red rings are and how inherently unpredictable startup outcomes are at the seed&nbsp;stage.</p><p>Some managers have done it for sure, but how much of it was luck? Any LP who has watched a manager&#8217;s performance dwindle from Fund I to Fund II to Fund III knows the answer to that question. Better to play the odds and bet on sound portfolio strategies over a few lucky&nbsp;GPs.</p><blockquote><p>What if a fund has a concentrated portfolio strategy with a massive reserve for follow-on investments? Can they reduce risk by concentrating capital in their big winners all the way to&nbsp;IPO?</p></blockquote><p>This is a strategy that looks great on paper but rarely works in the real world, because it relies on the following assumptions:</p><ol><li><p>The fund invests in big winners to start (unlikely with few investments)</p></li><li><p>The GPs know who their big winners are early&nbsp;on</p></li><li><p>The fund is able to keep re-investing in every growth&nbsp;round</p></li></ol><p>I&#8217;ve seen few examples of this working in practice, and even when it does, is it really a seed fund if most of the capital goes into growth rounds? There&#8217;s nothing wrong with growth funds of course, so long as they&#8217;re marketed at&nbsp;such.</p><blockquote><p>The bullseyes you shared are just sample portfolios. How can you reach conclusions based on just a few simulations?</p></blockquote><p>Believe me, the statistician in me hates a sample size of n=1. The proper way to test different portfolio strategies is by running Monte Carlo simulations across thousands of trials, which is exactly what we did to develop Rebel&#8217;s portfolio strategy, as explained in my <a href="https://link.medium.com/D6Ootw8LdJb">power law post</a>. For this post, I opted for understandability over robustness, though I tried to choose bullseyes that are &#8220;typical&#8221; for each portfolio strategy.</p><p>I suppose this is a good place to wrap up this post. If you&#8217;re an LP, hopefully you&#8217;ll think about the &#8220;bullseye&#8221; for every fund you diligence. If you&#8217;re a GP, perhaps this can help you think through the right portfolio strategy for your fund. If you&#8217;re a founder, you only get one &#8220;shot&#8221; so make it&nbsp;count!</p>]]></content:encoded></item><item><title><![CDATA[On the ‘new’ Y Combinator]]></title><description><![CDATA[Over the last few years, Y Combinator has gone through two major transformations at the same time: the AI revolution & Garry Tan returning to YC as President & CEO.]]></description><link>https://insights.rebelfund.vc/p/on-the-new-y-combinator-3c28e548896c</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-the-new-y-combinator-3c28e548896c</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Tue, 04 Nov 2025 18:07:48 GMT</pubDate><enclosure url="https://substack-post-media.s3.amazonaws.com/public/images/10fb6188-504b-4534-81db-95fbc45e1e7c_1024x328.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>Over the last few years, <a href="https://www.ycombinator.com/">Y Combinator</a> has gone through two major transformations at the same time: the AI revolution &amp; <a href="https://www.linkedin.com/in/garrytan/">Garry Tan</a> returning to YC as President &amp; CEO. These changes have had a huge impact on the profile of YC startups and founders, which will be the topic of this&nbsp;post.</p><p>Since <a href="https://www.rebelfund.vc/">Rebel Fund</a> maintains the world&#8217;s largest database of YC startups, founders, and outcomes outside of YC itself, we&#8217;re often the first to notice shifts in the composition of YC batches, and are best-positioned to predict their long-term implications for investors. We collect this data mostly to train our proprietary <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">Rebel Theorem 4.0</a> ML/AI model to predict YC startup success, but it&#8217;s also helpful to track and report on recent YC&nbsp;trends.</p><p>In this post, I&#8217;ll summarize some changes we&#8217;ve observed in YC batches recently under Garry&#8217;s leadership, and what these changes might mean for early-stage YC-focused investors like&nbsp;us.</p><p><strong>Founder Trends</strong></p><p>Since Garry took over in 2023, YC founders have gotten <em>much</em> younger. The average age of founders had been holding steady around ~29 years since 2015, but now it stands at ~26 years and&nbsp;falling.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!KsYa!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F72d9db24-9e05-43a3-a866-c2719ca61c36_1024x328.jpeg 424w, https://substackcdn.com/image/fetch/$s_!KsYa!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F72d9db24-9e05-43a3-a866-c2719ca61c36_1024x328.jpeg 848w, https://substackcdn.com/image/fetch/$s_!KsYa!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F72d9db24-9e05-43a3-a866-c2719ca61c36_1024x328.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!KsYa!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F72d9db24-9e05-43a3-a866-c2719ca61c36_1024x328.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!KsYa!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F72d9db24-9e05-43a3-a866-c2719ca61c36_1024x328.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/72d9db24-9e05-43a3-a866-c2719ca61c36_1024x328.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!KsYa!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F72d9db24-9e05-43a3-a866-c2719ca61c36_1024x328.jpeg 424w, https://substackcdn.com/image/fetch/$s_!KsYa!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F72d9db24-9e05-43a3-a866-c2719ca61c36_1024x328.jpeg 848w, https://substackcdn.com/image/fetch/$s_!KsYa!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F72d9db24-9e05-43a3-a866-c2719ca61c36_1024x328.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!KsYa!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F72d9db24-9e05-43a3-a866-c2719ca61c36_1024x328.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p>I think this is for two reasons: 1) Garry has generally been refocusing YC on what made it successful in the first place, and 2) whenever there is a major technology platform shift underway like AI, the balance of power between youthfulness and experience tends to shift in favor of the&nbsp;former.</p><p>The age cohorts of YC founders that have been growing the fastest in recent years are the youngest, and for the first time the past decade, there are more YC founders under the age of 25 years than&nbsp;over!</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!T92m!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4922eaa7-1ddd-4a94-a8af-a01bd9b6a2c0_1024x509.jpeg 424w, https://substackcdn.com/image/fetch/$s_!T92m!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4922eaa7-1ddd-4a94-a8af-a01bd9b6a2c0_1024x509.jpeg 848w, https://substackcdn.com/image/fetch/$s_!T92m!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4922eaa7-1ddd-4a94-a8af-a01bd9b6a2c0_1024x509.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!T92m!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4922eaa7-1ddd-4a94-a8af-a01bd9b6a2c0_1024x509.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!T92m!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4922eaa7-1ddd-4a94-a8af-a01bd9b6a2c0_1024x509.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/4922eaa7-1ddd-4a94-a8af-a01bd9b6a2c0_1024x509.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!T92m!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4922eaa7-1ddd-4a94-a8af-a01bd9b6a2c0_1024x509.jpeg 424w, https://substackcdn.com/image/fetch/$s_!T92m!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4922eaa7-1ddd-4a94-a8af-a01bd9b6a2c0_1024x509.jpeg 848w, https://substackcdn.com/image/fetch/$s_!T92m!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4922eaa7-1ddd-4a94-a8af-a01bd9b6a2c0_1024x509.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!T92m!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4922eaa7-1ddd-4a94-a8af-a01bd9b6a2c0_1024x509.jpeg 1456w" sizes="100vw"></picture><div></div></div></div></figure></div><p>YC founders are also nerdier. At Rebel, we use AI to divide founders into &#8220;technical&#8221; vs &#8220;non-technical&#8221; backgrounds, and the &#8220;technical&#8221; group has really taken over since 2023&#8202;&#8212;&#8202;another return to YC&#8217;s historical roots.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!ngDw!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0b7c70d4-658a-4974-9479-00cd43fd3aea_1024x507.jpeg 424w, https://substackcdn.com/image/fetch/$s_!ngDw!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0b7c70d4-658a-4974-9479-00cd43fd3aea_1024x507.jpeg 848w, https://substackcdn.com/image/fetch/$s_!ngDw!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0b7c70d4-658a-4974-9479-00cd43fd3aea_1024x507.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!ngDw!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0b7c70d4-658a-4974-9479-00cd43fd3aea_1024x507.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!ngDw!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0b7c70d4-658a-4974-9479-00cd43fd3aea_1024x507.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/0b7c70d4-658a-4974-9479-00cd43fd3aea_1024x507.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!ngDw!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0b7c70d4-658a-4974-9479-00cd43fd3aea_1024x507.jpeg 424w, https://substackcdn.com/image/fetch/$s_!ngDw!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0b7c70d4-658a-4974-9479-00cd43fd3aea_1024x507.jpeg 848w, https://substackcdn.com/image/fetch/$s_!ngDw!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0b7c70d4-658a-4974-9479-00cd43fd3aea_1024x507.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!ngDw!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0b7c70d4-658a-4974-9479-00cd43fd3aea_1024x507.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>We also look at how &#8220;pedigreed&#8221; YC founders are in various ways, and have observed that more YC founders graduated from top universities than ever in recent years. For the first time last year, <em>more than half</em> of YC founders graduated from a top 20 university.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!6D1s!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F34c3cc00-a5b4-43ec-8c64-27cf13673399_1024x706.jpeg 424w, https://substackcdn.com/image/fetch/$s_!6D1s!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F34c3cc00-a5b4-43ec-8c64-27cf13673399_1024x706.jpeg 848w, https://substackcdn.com/image/fetch/$s_!6D1s!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F34c3cc00-a5b4-43ec-8c64-27cf13673399_1024x706.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!6D1s!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F34c3cc00-a5b4-43ec-8c64-27cf13673399_1024x706.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!6D1s!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F34c3cc00-a5b4-43ec-8c64-27cf13673399_1024x706.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/34c3cc00-a5b4-43ec-8c64-27cf13673399_1024x706.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!6D1s!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F34c3cc00-a5b4-43ec-8c64-27cf13673399_1024x706.jpeg 424w, https://substackcdn.com/image/fetch/$s_!6D1s!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F34c3cc00-a5b4-43ec-8c64-27cf13673399_1024x706.jpeg 848w, https://substackcdn.com/image/fetch/$s_!6D1s!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F34c3cc00-a5b4-43ec-8c64-27cf13673399_1024x706.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!6D1s!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F34c3cc00-a5b4-43ec-8c64-27cf13673399_1024x706.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>There has been an even bigger jump in the proportion of YC founders with prior work experience at top employers (i.e., those most predictive of YC founder success). In recent years, nearly half of YC founders have worked at a top 20 employer vs only ~25% a decade&nbsp;ago.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!GS_I!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd029389-2ce3-4eab-988b-2e395eeac31c_1024x703.jpeg 424w, https://substackcdn.com/image/fetch/$s_!GS_I!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd029389-2ce3-4eab-988b-2e395eeac31c_1024x703.jpeg 848w, https://substackcdn.com/image/fetch/$s_!GS_I!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd029389-2ce3-4eab-988b-2e395eeac31c_1024x703.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!GS_I!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd029389-2ce3-4eab-988b-2e395eeac31c_1024x703.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!GS_I!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd029389-2ce3-4eab-988b-2e395eeac31c_1024x703.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/dd029389-2ce3-4eab-988b-2e395eeac31c_1024x703.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!GS_I!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd029389-2ce3-4eab-988b-2e395eeac31c_1024x703.jpeg 424w, https://substackcdn.com/image/fetch/$s_!GS_I!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd029389-2ce3-4eab-988b-2e395eeac31c_1024x703.jpeg 848w, https://substackcdn.com/image/fetch/$s_!GS_I!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd029389-2ce3-4eab-988b-2e395eeac31c_1024x703.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!GS_I!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdd029389-2ce3-4eab-988b-2e395eeac31c_1024x703.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>At Rebel, we use an AI model to infer the personality profile of each YC founder based on their career history and online footprint. Since Garry took the helm, average YC founder personality profiles have shifted, with much lower levels of Dominance and higher levels of Conscientiousness.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!hfnD!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7dfe593f-ae36-49b8-b9df-d73bb9232f02_1024x501.jpeg 424w, https://substackcdn.com/image/fetch/$s_!hfnD!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7dfe593f-ae36-49b8-b9df-d73bb9232f02_1024x501.jpeg 848w, https://substackcdn.com/image/fetch/$s_!hfnD!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7dfe593f-ae36-49b8-b9df-d73bb9232f02_1024x501.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!hfnD!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7dfe593f-ae36-49b8-b9df-d73bb9232f02_1024x501.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!hfnD!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7dfe593f-ae36-49b8-b9df-d73bb9232f02_1024x501.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/7dfe593f-ae36-49b8-b9df-d73bb9232f02_1024x501.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!hfnD!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7dfe593f-ae36-49b8-b9df-d73bb9232f02_1024x501.jpeg 424w, https://substackcdn.com/image/fetch/$s_!hfnD!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7dfe593f-ae36-49b8-b9df-d73bb9232f02_1024x501.jpeg 848w, https://substackcdn.com/image/fetch/$s_!hfnD!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7dfe593f-ae36-49b8-b9df-d73bb9232f02_1024x501.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!hfnD!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7dfe593f-ae36-49b8-b9df-d73bb9232f02_1024x501.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>We can also observe Garry&#8217;s bias towards community-building, with a massive leap in 2023 in the percentage of YC founders who had previously worked at a YC&nbsp;startup.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!4U3B!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab8f6173-0f48-4e15-abf6-3397f7f8b1a6_1024x402.jpeg 424w, https://substackcdn.com/image/fetch/$s_!4U3B!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab8f6173-0f48-4e15-abf6-3397f7f8b1a6_1024x402.jpeg 848w, https://substackcdn.com/image/fetch/$s_!4U3B!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab8f6173-0f48-4e15-abf6-3397f7f8b1a6_1024x402.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!4U3B!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab8f6173-0f48-4e15-abf6-3397f7f8b1a6_1024x402.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!4U3B!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab8f6173-0f48-4e15-abf6-3397f7f8b1a6_1024x402.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/ab8f6173-0f48-4e15-abf6-3397f7f8b1a6_1024x402.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!4U3B!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab8f6173-0f48-4e15-abf6-3397f7f8b1a6_1024x402.jpeg 424w, https://substackcdn.com/image/fetch/$s_!4U3B!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab8f6173-0f48-4e15-abf6-3397f7f8b1a6_1024x402.jpeg 848w, https://substackcdn.com/image/fetch/$s_!4U3B!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab8f6173-0f48-4e15-abf6-3397f7f8b1a6_1024x402.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!4U3B!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fab8f6173-0f48-4e15-abf6-3397f7f8b1a6_1024x402.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Last but not least, we can see Garry&#8217;s strong conviction in the San Francisco Bay&#8202;&#8212;&#8202;there has been a huge increase in the proportion of YC startups located in the Bay, now approaching 85%!</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!kGlZ!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F595c0142-de0e-4a87-89ad-7e14009a3460_1024x494.jpeg 424w, https://substackcdn.com/image/fetch/$s_!kGlZ!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F595c0142-de0e-4a87-89ad-7e14009a3460_1024x494.jpeg 848w, https://substackcdn.com/image/fetch/$s_!kGlZ!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F595c0142-de0e-4a87-89ad-7e14009a3460_1024x494.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!kGlZ!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F595c0142-de0e-4a87-89ad-7e14009a3460_1024x494.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!kGlZ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F595c0142-de0e-4a87-89ad-7e14009a3460_1024x494.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/595c0142-de0e-4a87-89ad-7e14009a3460_1024x494.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!kGlZ!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F595c0142-de0e-4a87-89ad-7e14009a3460_1024x494.jpeg 424w, https://substackcdn.com/image/fetch/$s_!kGlZ!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F595c0142-de0e-4a87-89ad-7e14009a3460_1024x494.jpeg 848w, https://substackcdn.com/image/fetch/$s_!kGlZ!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F595c0142-de0e-4a87-89ad-7e14009a3460_1024x494.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!kGlZ!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F595c0142-de0e-4a87-89ad-7e14009a3460_1024x494.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>So in short, YC has become &#8220;more YC&#8221; than ever, with a refocus on young, technical, pedigreed, Silicon Valley-based founders.</p><p><strong>Implications for YC Investors</strong></p><p>So what do these dramatic changes in YC founders and startups mean for early-stage YC-focused investors like Rebel? According to my data science team, it&#8217;s a mixed bag. Let&#8217;s start with the implications of founder&nbsp;age.</p><p>When we trained our <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">Rebel Theorem 4.0</a> ML/AI algorithm across millions of YC startup and founder data points, we found that characteristics related to founder age (years of work experience, years since university, etc) are some of the most powerful predictors of startup&nbsp;success.</p><p>We also found there is a &#8220;sweet spot&#8221; in founder age that properly balances the advantages and disadvantages of youth vs experience&#8202;&#8212;&#8202;<strong>26&#8211;30 years&nbsp;old</strong>.</p><p>Founders younger than 26 are much more likely to have &#8220;dead&#8221; startups, and founders older than 30 are much more likely to have &#8220;zombie&#8221; startups (still alive, but haven&#8217;t achieved breakout success). The odds of founding a &#8220;success&#8221; startup, which Rebel defines as a $60M+ valuation and still operating, really peaks in the 26&#8211;30 year old&nbsp;cohort.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!f7Y-!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F00c25daa-2cfc-45d6-9723-e9a15cc1132a_1024x344.jpeg 424w, https://substackcdn.com/image/fetch/$s_!f7Y-!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F00c25daa-2cfc-45d6-9723-e9a15cc1132a_1024x344.jpeg 848w, https://substackcdn.com/image/fetch/$s_!f7Y-!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F00c25daa-2cfc-45d6-9723-e9a15cc1132a_1024x344.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!f7Y-!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F00c25daa-2cfc-45d6-9723-e9a15cc1132a_1024x344.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!f7Y-!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F00c25daa-2cfc-45d6-9723-e9a15cc1132a_1024x344.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/00c25daa-2cfc-45d6-9723-e9a15cc1132a_1024x344.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!f7Y-!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F00c25daa-2cfc-45d6-9723-e9a15cc1132a_1024x344.jpeg 424w, https://substackcdn.com/image/fetch/$s_!f7Y-!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F00c25daa-2cfc-45d6-9723-e9a15cc1132a_1024x344.jpeg 848w, https://substackcdn.com/image/fetch/$s_!f7Y-!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F00c25daa-2cfc-45d6-9723-e9a15cc1132a_1024x344.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!f7Y-!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F00c25daa-2cfc-45d6-9723-e9a15cc1132a_1024x344.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>So, with over half of new YC founders now aged 25 and younger, you&#8217;d think there may be rocky times ahead, but founder age and its derivatives are just a few of the <em>hundreds</em> of features that <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3\">Rebel Theorem 4.0</a> uses to predict YC startup&nbsp;success.</p><p>For example, today&#8217;s younger YC founders score <em>higher</em> on 4 out of 6 <a href="https://jaredheyman.medium.com/on-founder-product-fit-7f1ad68d7af5">founder-product fit</a> metrics we track at Rebel, and these scores are also highly-predictive of successful YC startup outcomes.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Q1FG!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61c07718-b062-42fd-bb6d-5934879ad169_1024x422.png 424w, https://substackcdn.com/image/fetch/$s_!Q1FG!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61c07718-b062-42fd-bb6d-5934879ad169_1024x422.png 848w, https://substackcdn.com/image/fetch/$s_!Q1FG!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61c07718-b062-42fd-bb6d-5934879ad169_1024x422.png 1272w, https://substackcdn.com/image/fetch/$s_!Q1FG!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61c07718-b062-42fd-bb6d-5934879ad169_1024x422.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Q1FG!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61c07718-b062-42fd-bb6d-5934879ad169_1024x422.png" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/61c07718-b062-42fd-bb6d-5934879ad169_1024x422.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!Q1FG!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61c07718-b062-42fd-bb6d-5934879ad169_1024x422.png 424w, https://substackcdn.com/image/fetch/$s_!Q1FG!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61c07718-b062-42fd-bb6d-5934879ad169_1024x422.png 848w, https://substackcdn.com/image/fetch/$s_!Q1FG!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61c07718-b062-42fd-bb6d-5934879ad169_1024x422.png 1272w, https://substackcdn.com/image/fetch/$s_!Q1FG!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F61c07718-b062-42fd-bb6d-5934879ad169_1024x422.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>The shift towards more YC startups in the SF Bay Area is also a mixed bag, because while SF-based startups are significantly more likely to be successful than their non-SF peers, they&#8217;re also more likely to fail (Silicon Valley is &#8220;go big or go&nbsp;home&#8221;).</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!rD2W!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5fd4314-4bee-4f55-8f23-dc41dd566b2f_1024x495.jpeg 424w, https://substackcdn.com/image/fetch/$s_!rD2W!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5fd4314-4bee-4f55-8f23-dc41dd566b2f_1024x495.jpeg 848w, https://substackcdn.com/image/fetch/$s_!rD2W!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5fd4314-4bee-4f55-8f23-dc41dd566b2f_1024x495.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!rD2W!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5fd4314-4bee-4f55-8f23-dc41dd566b2f_1024x495.jpeg 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!rD2W!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5fd4314-4bee-4f55-8f23-dc41dd566b2f_1024x495.jpeg" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/e5fd4314-4bee-4f55-8f23-dc41dd566b2f_1024x495.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!rD2W!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5fd4314-4bee-4f55-8f23-dc41dd566b2f_1024x495.jpeg 424w, https://substackcdn.com/image/fetch/$s_!rD2W!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5fd4314-4bee-4f55-8f23-dc41dd566b2f_1024x495.jpeg 848w, https://substackcdn.com/image/fetch/$s_!rD2W!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5fd4314-4bee-4f55-8f23-dc41dd566b2f_1024x495.jpeg 1272w, https://substackcdn.com/image/fetch/$s_!rD2W!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe5fd4314-4bee-4f55-8f23-dc41dd566b2f_1024x495.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>At Rebel, we also look at hundreds of other aspects of founder pedigree, work experience, demographics, previous startups founded, <a href="https://jaredheyman.medium.com/on-the-5-archetypes-of-top-yc-founders-b0a81e8e09fc">founder archetypes</a>, and many more categories I haven&#8217;t yet written about publicly.</p><p>I could further dissect the new profile of YC startups under Garry&#8217;s leadership, but the bottom line is the &#8216;new&#8217; YC isn&#8217;t necessarily better or worse, just different. Only time will tell the net effect of YC&#8217;s latest selection criteria on startup outcomes, especially given how wildly different the technology world has become since the dawn of advanced&nbsp;AI.</p><p>One thing we know for certain is YC startups not only <em>look</em> different than before, but they also <em>feel</em> different. Founders are hungrier, there is a palpable sense of urgency, revenues are growing faster, and the stakes feel higher. In my opinion, it&#8217;s the best time ever to invest in YC startups.</p>]]></content:encoded></item><item><title><![CDATA[On the 5 archetypes of top YC founders]]></title><description><![CDATA[Rebel Fund has proudly invested in 500+ talented YC founders over the past several years across hundreds of startups.]]></description><link>https://insights.rebelfund.vc/p/on-the-5-archetypes-of-top-yc-founders-b0a81e8e09fc</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-the-5-archetypes-of-top-yc-founders-b0a81e8e09fc</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Thu, 21 Aug 2025 16:53:21 GMT</pubDate><enclosure url="https://cdn-images-1.medium.com/max/800/1*AXg75BmA44hT13WKVckB4w.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><a href="https://www.rebelfund.vc/">Rebel Fund</a> has proudly invested in 500+ talented YC founders over the past several years across hundreds of startups. Along the way, we&#8217;ve built the world&#8217;s most comprehensive database of YC startups and outcomes, now tracking millions of data points across every YC company and founder in history, in large part to train our proprietary <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">Rebel Theorem 4.0</a> ML/AI algorithm to accurately predict YC startup success.</p><p>Our algorithm weighs hundreds of YC company and founder characteristics or &#8216;features&#8217; across 100+ decision trees to ultimately assign a probabilty of success or failure to every new YC startup. It&#8217;s grown so complex that even our data and engineering team can&#8217;t fully explain its inner workings.</p><p>So, to demystify our model a bit, we simplified&#185; it into ~30 key features related to YC founders that give the model ~80% of its predictive power. We then performed a cluster analysis&#178; on these features to divide founders into 5 archetypes that are both human-understandable and predictive of startup outcomes. This blog post will reveal our top YC founder archetypes and share which ones are most likely to build $1B+ companies.</p><h3><strong>Archetype #1&#8202;&#8212;&#8202;&#8220;Charismatic Hustlers&#8221;</strong></h3><p><strong>Successful examples:</strong></p><p><a href="https://www.linkedin.com/in/kaarelkotkas/%5C">Kaarel Kotkas</a> @ Veriff (valuation ~$1.5B)</p><p><a href="https://www.linkedin.com/in/jake-loosararian-23981350/">Jake Loosarian</a> @ Gecko Robotics (valuation ~$1.6B)</p><p><strong>One-liner:</strong></p><blockquote><p>Charismatic, quick-moving sellers with minimal prior experience and weak technical/industry fit.</p></blockquote><p><strong>Key characteristics:</strong></p><p>These founders win attention and momentum through charisma, speed, persistence, and &#8220;hustling&#8221; into opportunities. They have shorter resumes than their YC peers, with just ~6 years work experience across few job roles, and rarely at top Silicon Valley employers. Academic backgrounds are less impressive than their peers, and their <a href="https://jaredheyman.medium.com/on-founder-product-fit-7f1ad68d7af5%5C">founder-product fit</a> scores are lower. They&#8217;re also less technical and more likely to be female than other archetypes.</p><p>They compensate for their less-than-stellar pedigrees with an outgoing personality and leadership style, showing very high levels of Dominance, Expressiveness, Social, and Pace traits, and the lowest levels of Risk Aversion, Skepticism, Leniency, and Pragmatism among all archetypes. This founder is fast, bold, inspiring, often light on analytics.</p><p><strong>Rebel Theorem outlook:</strong></p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*AXg75BmA44hT13WKVckB4w.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*AXg75BmA44hT13WKVckB4w.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*AXg75BmA44hT13WKVckB4w.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*AXg75BmA44hT13WKVckB4w.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*AXg75BmA44hT13WKVckB4w.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*AXg75BmA44hT13WKVckB4w.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*AXg75BmA44hT13WKVckB4w.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*AXg75BmA44hT13WKVckB4w.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*AXg75BmA44hT13WKVckB4w.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*AXg75BmA44hT13WKVckB4w.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Our &#8220;Charimatic Hustlers&#8221; are less likely to build successful companies than most other archetypes, but are also less likely to stall out or fail. They can keep their startup alive for years through sheer force of personality, but their companies don&#8217;t often become unicorns.</p><p>Their natural charisma makes them excellent at startup pitches and raising capital, but investors should be cautious with them.</p><h3>Archetype #2&#8202;&#8212;&#8202;&#8220;<strong>Bay-Area Value Builders</strong>&#8221;</h3><p><strong>Successful examples:</strong></p><p><a href="https://www.linkedin.com/in/parkerconrad/">Parker Conrad</a> @ Rippling (valuation ~$19B)</p><p><a href="https://www.linkedin.com/in/lakssrini/%5C">Laks Srini</a> @ ZeroDown (valuation ~$190M)</p><p>(Both my YC batchmates while @ Zenefits. Go W13!)</p><p><strong>One-liner:</strong></p><blockquote><p>Experienced and well&#8209;connected, with credible signals&#8202;&#8212;&#8202;the group most likely to produce breakout successes.</p></blockquote><p><strong>Key characteristics:</strong></p><p>Their experience, networks, and target markets skew toward the Bay Area, and their track records point to building meaningful company value. They have above&#8209;average work experience (~10 years) prior to founding, more than 5 years prior experience as co-founders, and more prior job roles and employers than average. They often achieved strong prior startup outcomes as well, with large teams managed, higher prior startup valuations, more funding raised as co-founders than their peers.</p><p>The also have a strong Bay Area and North America footprint, with the longest tenure working in San Francisco. They&#8217;re most likely building a B2B company that sells to other startups. These founders&#8217; personality skews moderately assertive and action-oriented, exhibiting slightly lower Leniency than their peers, and variability around Skepticism and Sociability. They are the classic startup founder profile that many VCs gravitate towards, and for good reason.</p><p><strong>Rebel Theorem outlook:</strong></p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*Pm6wQ8p5ViKT9jPiDZTwPg.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Pm6wQ8p5ViKT9jPiDZTwPg.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Pm6wQ8p5ViKT9jPiDZTwPg.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Pm6wQ8p5ViKT9jPiDZTwPg.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*Pm6wQ8p5ViKT9jPiDZTwPg.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*Pm6wQ8p5ViKT9jPiDZTwPg.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*Pm6wQ8p5ViKT9jPiDZTwPg.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Pm6wQ8p5ViKT9jPiDZTwPg.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Pm6wQ8p5ViKT9jPiDZTwPg.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Pm6wQ8p5ViKT9jPiDZTwPg.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Our &#8220;Bay-Area Value Builders&#8221; are much more likely to build successful startups than their peers, though with high variability. They&#8217;re also much less likely to build &#8216;zombie&#8217; startups that neither achieve stellar outcomes nor die. Their startups are less likely to fail than most other archetypes&#8217;, but again with a high variability.</p><p>You can think of these founders as a growth stock in human form&#8202;&#8212;&#8202;very high success potential but with higher beta. Investors love them and their rounds often get <a href="https://medium.com/p/on-investing-in-hot-y-combinator-startups-aed26869f77c">hot at YC Demo Day</a>.</p><h3>Archetype #3&#8202;&#8212;&#8202;&#8220;<strong>Pedigreed B2B Operators</strong>&#8221;</h3><p><strong>Successful examples:</strong></p><p><a href="https://www.linkedin.com/in/erik-goldman/">Erik Goldman</a> @ Vanta (valuation ~$4B)</p><p><a href="https://www.linkedin.com/in/jimbru/">Jim Brusstar</a> @ Treasury Prime (valuation ~$340M)</p><p><strong>One-liner:</strong></p><blockquote><p>High-signal resumes building B2B products, with solid (not top) success odds.</p></blockquote><p><strong>Key characteristics:</strong></p><p>These highly pedigreed founders combine top university degrees and prior employers with a builder-mindset and operational discipline. They have above-average prior work experience (~9 years), often across a higher number of roles and companies than their peers. They have little prior co-founder experience, and when they did build a startup in the past, they only led small teams and did not achieve breakout success. They probably left university after a bachelor degree, and 75% of them have technical backgrounds. They&#8217;re most likely building a B2B software startup in the Bay Area.</p><p>Personality-wise, these founders are measured, thoughtful decision-makers who favor stability and careful planning over speed or visibility. With lower Dominance, Expressiveness, Social, and Pace, they tend to lead quietly, avoid the spotlight, and take time before acting. Their higher Leniency, Pragmatism, Skepticism, and Risk Aversion make them accommodating yet analytical, balancing openness to others&#8217; perspectives with a critical mindset and a strong preference for minimizing risk.</p><p><strong>Rebel Theorem outlook:</strong></p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*Fi4ILs4xxda5gENHBa9I0g.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Fi4ILs4xxda5gENHBa9I0g.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Fi4ILs4xxda5gENHBa9I0g.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Fi4ILs4xxda5gENHBa9I0g.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*Fi4ILs4xxda5gENHBa9I0g.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*Fi4ILs4xxda5gENHBa9I0g.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*Fi4ILs4xxda5gENHBa9I0g.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Fi4ILs4xxda5gENHBa9I0g.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Fi4ILs4xxda5gENHBa9I0g.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Fi4ILs4xxda5gENHBa9I0g.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Our &#8220;Pedigreed B2B Operators&#8221; have success odds second only to the &#8220;Bay-Area Value Builders&#8221; but with less variability. Their zombie odds are comparable to the prior group, but again with less variability, and their odds of failure are a bit higher than average.</p><p>This archetype is often overlooked by investors because they&#8217;re not as charismatic or proven as other founders, but they are solid bets statistically, and top VCs (and ML algorithms!) recognize this.</p><h3>Archetype #4&#8202;&#8212;&#8202;&#8220;Veteran Generalists&#8221;</h3><p><strong>Successful examples:</strong></p><p><a href="https://www.linkedin.com/in/idrismokhtarzada/">Idris Mokhtarzada</a> @ Truebill (valuation ~$1.2B)</p><p><a href="https://www.linkedin.com/in/renesaul/">Rene Saul</a> @ Kapital Bank (valuation ~$800M)</p><p><strong>One-liner:</strong></p><blockquote><p>Seasoned, versatile operators with real execution mileage and low pedigree; durable but not explosive.</p></blockquote><p><strong>Key characteristics:</strong></p><p>These founders have extensive, varied experience across roles and companies, with the highest overall work experience prior to founding (~12 years) and prior co-founder experience (~6 years). They also have the highest average number of work experiences, usually across several employers. As a result of their wide work experience, they achieve the highest <a href="https://jaredheyman.medium.com/on-founder-product-fit-7f1ad68d7af5">founder-product fit</a> scores.</p><p>They have significantly more prior experience scaling startups than their peers, with larger teams managed and higher valuations achieved. They&#8217;re less pedigreed, with work backgrounds at companies ranked outside the top-tier, and likely didn&#8217;t attend elite universities, but are more likely to have earned an advanced degree. Their Bay Area experience is the lowest of all archetypes, and they have a sector bias towards fintech. They also have the highest proportion of male founders across all groups.</p><p>Their personality mix includes above-average Dominance, Expresiveness and Pace, but lower Leniency, Risk Aversion, Pragmatism and Skepticism traits. This founder is naturally assertive, communicative, and energetic&#8202;&#8212;&#8202;someone who likes to move fast, speak up, and take charge when needed. They don&#8217;t overanalyze, and are driven more by gut and ambition than careful planning or cautious evaluation.</p><p><strong>Rebel Theorem outlook:</strong></p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*Ew11SQlARyk7XrMRiupQGA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Ew11SQlARyk7XrMRiupQGA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Ew11SQlARyk7XrMRiupQGA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Ew11SQlARyk7XrMRiupQGA.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*Ew11SQlARyk7XrMRiupQGA.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*Ew11SQlARyk7XrMRiupQGA.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*Ew11SQlARyk7XrMRiupQGA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Ew11SQlARyk7XrMRiupQGA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Ew11SQlARyk7XrMRiupQGA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Ew11SQlARyk7XrMRiupQGA.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Our &#8220;Veteran Generalists&#8221; have slighly above average odds of success, but interestingly, <em>much</em> lower odds of failure than their peers. The balance goes to the zombie category, meaning their companies continue to operate but never achieve a very high valuation.</p><p>Investors often think of these founders as safe bets, but that&#8217;s only correct in the sense their startups rarely crash and burn, thanks to their vast experience and maturity. However, traditional venture funds rely on super-outlier &#8216;unicorns&#8217; in their portfoliio to achieve high returns, and &#8220;Veteran Generalists&#8221; are less likely to build them them than our last two archetypes, so most VCs should avoid them.</p><h3>Archetype #5&#8202;&#8212;&#8202;&#8220;Pragmatic Engineers&#8221;</h3><p><strong>Successful examples:</strong></p><p><a href="https://www.linkedin.com/in/fangsterr/">Andy Fang</a> @ Doordash (valuation ~$100B)</p><p><a href="https://www.linkedin.com/in/joe-passanante/">Joe Passanante</a> @ Arist (valuation ~$60M)</p><p><strong>One-liner:</strong></p><blockquote><p>Methodical founders who avoid big risks, often leading to small-scale outcomes.</p></blockquote><p><strong>Key characteristics:</strong></p><p>These are highly technical, cautious founders who prioritize practical decision-making over bold moves. They have the most technical profile among our archetypes and the highest education-fit score. They&#8217;re young, with only ~6 years of work experience on average, and have the least co-founder and people management experience. However, they have an above-average university pedigree, and often attended the same university as their co-founders. They also have the lowest <a href="https://jaredheyman.medium.com/on-founder-product-fit-7f1ad68d7af5">founder-product fit</a> scores in terms of work experience vs go-to-market, which means their backgrounds may be misaligned with their startup&#8217;s GTM/growth needs.</p><p>The personalities of these founders include high Pragmatism, Skepticism, Risk Aversion and Leniency, pared with very low Expressiveness, Dominance, Social &amp; Pace&#8202;&#8212;&#8202;think the &#8220;nerdy engineer&#8221; stereotype. This person is a cautious and pragmatic thinker who prefers to operate behind the scenes, avoiding conflict and fast-paced environments. Since they&#8217;re highly skeptical and risk-averse, they prioritize careful planning and low-risk decisions, while showing flexibility and openness to others&#8217; input.</p><p><strong>Rebel Theorem outlook:</strong></p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*GYCpiJCleI6ivHHXSxJnBQ.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*GYCpiJCleI6ivHHXSxJnBQ.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*GYCpiJCleI6ivHHXSxJnBQ.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*GYCpiJCleI6ivHHXSxJnBQ.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*GYCpiJCleI6ivHHXSxJnBQ.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*GYCpiJCleI6ivHHXSxJnBQ.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*GYCpiJCleI6ivHHXSxJnBQ.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*GYCpiJCleI6ivHHXSxJnBQ.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*GYCpiJCleI6ivHHXSxJnBQ.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*GYCpiJCleI6ivHHXSxJnBQ.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>YC has long held a bias towards highly-technical founders, so our &#8220;Pragmatic Engineers&#8221; are a common feature in YC batches. However, they are the founder archetype that&#8217;s <em>least</em> likely to achieve a successful startup outcome and the most likely to have a failed startup, each within a low variance&#8202;&#8212;&#8202;at least on their own.</p><p>The good news for our &#8220;Pragmatic Engineers&#8221; is founding a YC startup is often a team sport&#8202;&#8212;&#8202;the vast majority of YC startups have 2&#8211;3 co-founders. If a &#8220;Pragmatic Engineer&#8221; takes a CTO role and works with another archetype in a CEO role, their company&#8217;s odds of success goes up dramatically.</p><p>This leads me to a few other caveats I should make before wrapping up this post:</p><ol><li><p><strong>Founder archetypes are messy</strong>&#8202;&#8212;&#8202;People and personalites are complex, and most YC founders don&#8217;t fall 100% within a certain archetype&#179;. This is just a convenient mental model for dividing founders into &#8220;types&#8221;.</p></li><li><p><strong>Archetypes aren&#8217;t destiny&#8202;</strong>&#8212;&#8202;Every founder archetype produced many successful, failed, and zombie startups. In fact, some of our more common but lower-performing archetypes in terms of success odds actually produced <em>more</em> unicorn founders in absolute number.</p></li><li><p><strong>Startups are multi-faceted</strong>&#8202;&#8212;&#8202;Much more goes into startup investment decisions (and <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">Rebel Theorem 4.0</a> scores) than founder archetypes. A good investor also considers product, market, traction, defensibility, timing, and more.</p></li></ol><p>If you&#8217;re an investor, I hope you found this post helpful for reframing how you think about which YC founders to bet on.</p><p>If you&#8217;re a founder, which archetype are you???</p><p>Technical footnotes:</p><p>&#185;To reduce model complexity for the sake of explainability, we applied Principal Component Analysis (PCA)&#8202;&#8212;&#8202;a method that condenses many correlated variables into fewer &#8220;components&#8221; that preserve most of the information.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*Sg6RghKYCnYl2PJPpws_2Q.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Sg6RghKYCnYl2PJPpws_2Q.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Sg6RghKYCnYl2PJPpws_2Q.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Sg6RghKYCnYl2PJPpws_2Q.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*Sg6RghKYCnYl2PJPpws_2Q.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*Sg6RghKYCnYl2PJPpws_2Q.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*Sg6RghKYCnYl2PJPpws_2Q.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Sg6RghKYCnYl2PJPpws_2Q.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Sg6RghKYCnYl2PJPpws_2Q.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Sg6RghKYCnYl2PJPpws_2Q.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>&#178;We used K-Means, an algorithm that groups observations so each is assigned to the nearest &#8220;cluster center.&#8221; Experiments with different numbers of clusters showed that 5 clusters provided the best balance between interpretability and separation. We ran K-Means multiple times with different starting points (&#8220;seeds&#8221;) and measured consistency using the Adjusted Rand Index (ARI)&#8202;&#8212;&#8202;a metric that compares how similar two different clusterings are, adjusting for chance agreement. ARI ranges from 0 (no better than random) to 1 (perfectly identical assignments). Our mean ARI was 0.9874, showing almost identical assignments across runs&#8202;&#8212;&#8202;indicating robust and reliable cluster definitions.</p><p>&#179;We calculated the Silhouette Score, which measures how well each observation fits within its assigned cluster compared to other clusters. It ranges from -1 (poor clustering&#8202;&#8212;&#8202;points may be in the wrong cluster) to 1 (perfectly separated clusters), with scores near 0 indicating overlapping boundaries. Our result suggests that our clusters are distinct but not widely separated&#8202;&#8212;&#8202;typical for high-dimensional human data.</p>]]></content:encoded></item><item><title><![CDATA[On 9 data-driven tips for YC startup founders]]></title><description><![CDATA[Since Rebel Fund invests exclusively in seed-stage Y Combinator startups, the dozens of blog posts I&#8217;ve published over the years are focused mostly on how investors like us can improve their odds of investing in tomorrow&#8217;s YC unicorns.]]></description><link>https://insights.rebelfund.vc/p/on-9-data-driven-tips-for-yc-startup-founders-432908b40333</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-9-data-driven-tips-for-yc-startup-founders-432908b40333</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Thu, 24 Jul 2025 19:42:05 GMT</pubDate><enclosure url="https://cdn-images-1.medium.com/max/800/1*AtCh4q36R7290kJ9R7qMHA.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>Since <a href="https://www.rebelfund.vc/">Rebel Fund</a> invests exclusively in seed-stage <a href="https://www.ycombinator.com/">Y Combinator</a> startups, the <a href="https://jaredheyman.medium.com/">dozens of blog posts</a> I&#8217;ve published over the years are focused mostly on how investors like us can improve their odds of investing in tomorrow&#8217;s YC unicorns. However, this post will take a different approach&#8202;&#8212;&#8202;I&#8217;ll share some data-driven tips that <em><strong>YC founders</strong></em> can follow to maximize their odds of success based on what we&#8217;ve learned over 5+ years building the world&#8217;s most sophisticated <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">ML/AI algorithm</a> for predicting YC startup success.</p><p><strong>Tip #1&#8202;&#8212;&#8202;Swing for the fences</strong></p><p>Startup outcomes follow a <a href="https://link.medium.com/D6Ootw8LdJb">very steep power law curve</a>, such that ~6% of YC startups represent a whopping ~90% of the total valuation growth. The vast majority of startups enjoy no significant valuation growth at all, with a few big winners like Airbnb, Stripe, DoorDash, etc dwarfing even other decacorns ($10B+ valuations), which in turn dwarf even other unicorns ($1B+ valuations), which in turn dwarf the minicorns ($100M+ valuations).</p><p>Startup outcomes are relatively binary&#8212; either they&#8217;re a huge success or a failure. The implication is that founders should 1) pursue opportunities with massive upside potential rather than incremental gains, and 2) give it their all. The years after you start your first venture-backed startup will probably be the most consequential of your entire career.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*AtCh4q36R7290kJ9R7qMHA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*AtCh4q36R7290kJ9R7qMHA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*AtCh4q36R7290kJ9R7qMHA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*AtCh4q36R7290kJ9R7qMHA.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*AtCh4q36R7290kJ9R7qMHA.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*AtCh4q36R7290kJ9R7qMHA.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*AtCh4q36R7290kJ9R7qMHA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*AtCh4q36R7290kJ9R7qMHA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*AtCh4q36R7290kJ9R7qMHA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*AtCh4q36R7290kJ9R7qMHA.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p><strong>Tip #2&#8202;&#8212;&#8202;Be patient</strong></p><p>The median time for a YC startup to <a href="https://jaredheyman.medium.com/on-yc-startup-exits-2025-update-c6017e8e526e">achieve an exit</a> (acquisition or IPO) is about 3 years, but larger exits ($100M-$999M) take closer to 6 years, and unicorn exits ($1B+) typically take <em>nearly a decade</em>. Building a successful technology startup is a marathon, not a sprint.</p><p>You may see headlines about tech companies achieving ridiculous valuations ridiculously fast, but the reason they make headlines is they&#8217;re the exception rather than the rule. In the vast majority of cases, building a unicorn is a slow, painful, and unglamorous decade-long commitment.</p><p>Of course there are many exciting milestones along the way, but you should think of them as islands in a sea of hard work, relentless focus, and near-death experiences. If you don&#8217;t enjoy the journey of &#8220;building something people want&#8221; with its long hours and even longer odds, then you&#8217;ll never make it to the destination.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*TgwsYq_sLcrPrm0S8m9O9Q.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*TgwsYq_sLcrPrm0S8m9O9Q.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*TgwsYq_sLcrPrm0S8m9O9Q.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*TgwsYq_sLcrPrm0S8m9O9Q.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*TgwsYq_sLcrPrm0S8m9O9Q.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*TgwsYq_sLcrPrm0S8m9O9Q.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*TgwsYq_sLcrPrm0S8m9O9Q.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*TgwsYq_sLcrPrm0S8m9O9Q.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*TgwsYq_sLcrPrm0S8m9O9Q.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*TgwsYq_sLcrPrm0S8m9O9Q.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p><strong>Tip #3&#8202;&#8212;&#8202;Start young</strong></p><p>We found that the average <a href="https://jaredheyman.medium.com/on-yc-unicorn-founders-6530aad4bb89">YC unicorn founder</a> had 8 years work experience when they started their YC startup. Assuming they got their first job after graduating college at ~22 years of age, that means they were ~30 years old. Plenty of founders had more or less than this average, and a surprising number were only a few years into their career&#8202;&#8212;&#8202;though few had over 15 years experience.</p><p>Building a startup takes a lot of time and energy, and risk, so it&#8217;s better to start early in your career. Clearly some real-world work experience helps, but the trick is to properly balance energy and wisdom. Anecdotally, we&#8217;ve noticed that the current generation of &#8220;AI first&#8221; startup founders are younger than the historical average, probably because advanced AI is so fresh that relevant industry experience isn&#8217;t really to be had.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*4e03a8gGywXMAveypeLJUA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*4e03a8gGywXMAveypeLJUA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*4e03a8gGywXMAveypeLJUA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*4e03a8gGywXMAveypeLJUA.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*4e03a8gGywXMAveypeLJUA.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*4e03a8gGywXMAveypeLJUA.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*4e03a8gGywXMAveypeLJUA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*4e03a8gGywXMAveypeLJUA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*4e03a8gGywXMAveypeLJUA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*4e03a8gGywXMAveypeLJUA.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p><strong>Tip #4&#8202;&#8212;&#8202;Get co-founders</strong></p><p>One of the many things we learned training our <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">Rebel Theorem 4.0</a> ML/AI algorithm for predicting YC startup success is that the number of co-founders in a startup is positively correlated with successful outcomes. It&#8217;s not true in the absolute (you can have too many cooks in the kitchen) but partnering with another co-founder or two is a smart idea.</p><p>We&#8217;re working on a set of new algorithm features now that dissect exactly what characteristics of &#8220;co-founder fit&#8221; predict startup success, so I&#8217;ll have more to say on that soon. One thing we know for sure though is co-founder breakups are the #1 cause of early-stage startup failure. So at a minimum, make sure you know your co-founders well and you&#8217;re aligned in terms of long-term vision and values.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*jBgHvQwwvz_AbMQCXDNUKg.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*jBgHvQwwvz_AbMQCXDNUKg.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*jBgHvQwwvz_AbMQCXDNUKg.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*jBgHvQwwvz_AbMQCXDNUKg.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*jBgHvQwwvz_AbMQCXDNUKg.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*jBgHvQwwvz_AbMQCXDNUKg.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*jBgHvQwwvz_AbMQCXDNUKg.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*jBgHvQwwvz_AbMQCXDNUKg.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*jBgHvQwwvz_AbMQCXDNUKg.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*jBgHvQwwvz_AbMQCXDNUKg.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p><strong>Tip #5&#8202;&#8212;&#8202;Go to a top (technical) university</strong></p><p>Another thing we learned about <a href="https://jaredheyman.medium.com/on-yc-unicorn-founders-6530aad4bb89%5C">YC unicorn founders</a> is they often graduated from top-ranked universities with strong engineering programs (think Stanford or MIT).</p><p>There is a long tail of other universities represented amongst unicorn founders, so going to a top school really falls into the &#8220;helpful but not necessary&#8221; category, but having co-founders with strong technical and/or product backgrounds is vital based on our data.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*tHRdgu1XrRM7Sq_cLmBAFA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*tHRdgu1XrRM7Sq_cLmBAFA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*tHRdgu1XrRM7Sq_cLmBAFA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*tHRdgu1XrRM7Sq_cLmBAFA.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*tHRdgu1XrRM7Sq_cLmBAFA.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*tHRdgu1XrRM7Sq_cLmBAFA.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*tHRdgu1XrRM7Sq_cLmBAFA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*tHRdgu1XrRM7Sq_cLmBAFA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*tHRdgu1XrRM7Sq_cLmBAFA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*tHRdgu1XrRM7Sq_cLmBAFA.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p><strong>Tip #6&#8202;&#8212;&#8202;Cultivate key co-founder personality traits</strong></p><p><a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">Our algorithm</a> for predicting YC startup success is trained on inferred founder personality traits via a specialized AI model based on certain aspects of their online footprint. We&#8217;ve observed that certain key personality profiles are predictive of becoming a YC unicorn founder:</p><p><strong>High Dominance</strong>&#8202;&#8212;&#8202;results-oriented, confident, and decisive founders, preferring to take control of situations, focus on achieving goals, and direct in their communication</p><p><strong>High Conscientiousness</strong>&#8202;&#8212;&#8202;founders who place importance on accuracy, quality, and detailed work, with a strong sense of responsibility and a preference for working within established structures or guidelines</p><p><strong>Low Influence&#8202;&#8212;&#8202;</strong>reserved when it comes to social interactions, preferring to work independently or in smaller groups, don&#8217;t actively seek to be the center of attention, and prefer to rely on facts and logic rather than charm</p><p><strong>Low Steadiness</strong>&#8202;&#8212;&#8202;thrive in environments that are fast-paced and dynamic rather than stable or predictable, open to change and comfortable with a certain level of uncertainty, preferring variety over routine</p><p>Past a certain age you can&#8217;t really change your personality, but you can and should partner with co-founders who have these personality traits if you don&#8217;t yourself.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*NKJPGN7MqS2umuFEmXo3-w.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*NKJPGN7MqS2umuFEmXo3-w.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*NKJPGN7MqS2umuFEmXo3-w.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*NKJPGN7MqS2umuFEmXo3-w.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*NKJPGN7MqS2umuFEmXo3-w.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*NKJPGN7MqS2umuFEmXo3-w.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*NKJPGN7MqS2umuFEmXo3-w.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*NKJPGN7MqS2umuFEmXo3-w.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*NKJPGN7MqS2umuFEmXo3-w.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*NKJPGN7MqS2umuFEmXo3-w.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p><strong>Tip #7&#8202;&#8212;&#8202;Move to Silicon Valley</strong></p><p>Out of hundreds of YC founder and company characteristics we&#8217;ve found to be predictive of startup success via our <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">algorithm training</a>, simply whether or not the company is headquartered in San Francisco made the top 25! The vast majority of YC unicorns so far are based in the Bay Area.</p><p>There are many reasons for this which I&#8217;ll save for another post, but the data is clear&#8202;&#8212;&#8202;if you&#8217;re a founder, get your butt to Silicon Valley. A YC partner once described building a company in the Valley as &#8220;easy mode&#8221; and having built startups myself both inside and outside of the Valley, I can vouch for that.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*XRNd26JQPLurdK1zYoMClA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*XRNd26JQPLurdK1zYoMClA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*XRNd26JQPLurdK1zYoMClA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*XRNd26JQPLurdK1zYoMClA.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*XRNd26JQPLurdK1zYoMClA.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*XRNd26JQPLurdK1zYoMClA.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*XRNd26JQPLurdK1zYoMClA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*XRNd26JQPLurdK1zYoMClA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*XRNd26JQPLurdK1zYoMClA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*XRNd26JQPLurdK1zYoMClA.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p><strong>Tip #8&#8202;&#8212;&#8202;Assure founder-product fit</strong></p><p>You probably know the importance of product-market fit for startups, but our data shows that <a href="https://jaredheyman.medium.com/on-founder-product-fit-7f1ad68d7af5%5C">founder-product fit</a> is just as important.</p><p>At Rebel, we analyzed several metrics related to founder-product fit (i.e., how well founders&#8217; backgrounds align with the product they&#8217;re building) and discovered that while strong founder-product fit doesn&#8217;t necessarily predict success, a <em>lack of founder-product fit</em> is a strong predictor of failure.</p><p>It&#8217;s particularly dangerous when companies have poor founder-product fit around <strong>geography</strong> (i.e., founders don&#8217;t have backgrounds in the same geography of their target market), <strong>go-to market</strong> (i.e., founders don&#8217;t understand the startup&#8217;s go-to-market approach) and <strong>work experience</strong> (i.e., founders&#8217; prior work experience isn&#8217;t relevant to the company&#8217;s product).</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*L5qP7WKD0lvnUwOLQoY5bg.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*L5qP7WKD0lvnUwOLQoY5bg.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*L5qP7WKD0lvnUwOLQoY5bg.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*L5qP7WKD0lvnUwOLQoY5bg.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*L5qP7WKD0lvnUwOLQoY5bg.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*L5qP7WKD0lvnUwOLQoY5bg.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*L5qP7WKD0lvnUwOLQoY5bg.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*L5qP7WKD0lvnUwOLQoY5bg.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*L5qP7WKD0lvnUwOLQoY5bg.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*L5qP7WKD0lvnUwOLQoY5bg.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p><strong>Tip #9&#8202;&#8212;&#8202;Don&#8217;t worry if you&#8217;re not &#8220;hot&#8221; at Demo Day</strong></p><p>We analyzed whether companies that are <a href="https://medium.com/p/on-investing-in-hot-y-combinator-startups-aed26869f77c">&#8220;hot&#8221; at Demo Day</a> ended up having better outcomes than their peers. The bottom line is while there is a decent positive correlation between the total amount of capital a company raises at Demo Day and their long-term valuation growth, more than 90% of a company&#8217;s outcome is explained by other factors.</p><p>For founders, this means while it&#8217;s generally a positive signal if investors get excited about your company at Demo Day, a lack of excitement is far from a death sentence&#8202;&#8212;&#8202;your eventual success or failure will overwhelmingly be determined by other factors.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*saypLFt5K39QS_wODphkbQ.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*saypLFt5K39QS_wODphkbQ.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*saypLFt5K39QS_wODphkbQ.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*saypLFt5K39QS_wODphkbQ.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*saypLFt5K39QS_wODphkbQ.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*saypLFt5K39QS_wODphkbQ.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*saypLFt5K39QS_wODphkbQ.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*saypLFt5K39QS_wODphkbQ.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*saypLFt5K39QS_wODphkbQ.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*saypLFt5K39QS_wODphkbQ.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>I could pull many other tips for YC founders based on Rebel&#8217;s data, but I think this is a good place to stop for now. I hope some founders find this helpful. Now get back to talking to customers and building product!</p>]]></content:encoded></item><item><title><![CDATA[On founder-product fit]]></title><description><![CDATA[I&#8217;ve written dozens of blog posts over the years on how Rebel Fund uses data and algorithms to predict tomorrow&#8217;s most successful Y Combinator startups, and ultimately achieve an outsized return for our investors.]]></description><link>https://insights.rebelfund.vc/p/on-founder-product-fit-7f1ad68d7af5</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-founder-product-fit-7f1ad68d7af5</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Thu, 17 Jul 2025 16:20:52 GMT</pubDate><enclosure url="https://cdn-images-1.medium.com/max/800/1*hs2clNl0mb3IrTij4ssbDg@2x.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>I&#8217;ve written <a href="https://jaredheyman.medium.com">dozens of blog posts</a> over the years on how <a href="https://www.rebelfund.vc/">Rebel Fund</a> uses data and algorithms to predict tomorrow&#8217;s most successful Y Combinator startups, and ultimately achieve an outsized return for our investors.</p><p>My last post <a href="https://jaredheyman.medium.com/on-rebel-theorem-4-0-55d04b0732e3">On Rebel Theorem 4.0</a> unveils the latest release of our proprietary ML/AI model, which after 5+ years of development and millions of R&amp;D dollars invested, has proven incredibly powerful at predicting future YC unicorns.</p><p>I&#8217;m often asked by other investors how they can better leverage their own internal data to make better investment decisions. Telling the full story of how we developed Rebel Theorem would take too long, so I&#8217;ll focus this post on how we recently added a set of new features into our model around &#8220;founder-product fit&#8221; because I think it&#8217;s a perfect case study on how to brainstorm, test, and incorporate new features into an ML/AI model.</p><p>For context, our latest Rebel Theorem 4.0 model already includes 200+ features that we&#8217;ve developed, tested, validated, and incorporated over the years, but we&#8217;re always looking for new ways further enhance the model&#8217;s predictive accuracy. Now that we have advanced AI tools at our fingertips, we can be more creative with this than ever. Here are the steps we followed to incorporate some new founder-product fit features into Rebel Theorem.</p><p><strong>Step #1&#8202;&#8212; Brainstorm Features</strong></p><p>Our data science team started with a simple question: <em>Is the &#8220;fit&#8221; between a founder and their company predictive of startup outcomes?</em></p><p>On its face, it seems that it should be, but given the incredibly high stakes of seed-stage startup investing, we never add new features to our model without thoroughly backtesting them, no matter how intuitively predictive they may seem. So, we broke down this broad question into 3 categories with 8 testable questions across them.</p><p>Domain &amp; Industry Expertise:</p><p><em>1) Does the founder have any significant experience in this industry?</em></p><p><em>2) How relevant is the founder&#8217;s prior work experience to the company they are founding?</em></p><p><em>3) How relevant is the founder&#8217;s academic background to the startup?</em></p><p>Customer &amp; Market Insight:</p><p><em>4) How aligned are the geographies in which the founder has previously operated with the startup&#8217;s target market locations?</em></p><p><em>5) How well does the founder understand the current startup&#8217;s go-to-market approach (e.g., SaaS, marketplace, DTC)?</em></p><p><em>6) Has the founder previously worked with or applied the core technology driving their current startup, such as AI, blockchain, or CRISPR?</em></p><p>Leadership &amp; Organizational Readiness:</p><p><em>7) How clearly does the founder&#8217;s career path show progression toward larger scope and higher leadership responsibility?</em></p><p><em>8) How well has the founder demonstrated adaptability in moving between large-corporate environments and scrappy startup settings?</em></p><p><strong>Step #2&#8202;&#8212; Prompt AI Model</strong></p><p>Next we fine-tuned and prompted an AI model (OpenAI&#8217;s o3 model in this case) with lots of internal data we&#8217;ve collected on every YC company and founder in history, so it could score each founder-product pair on their fit across these 8 questions. Even though we&#8217;re testing qualitative features, we ultimately need numerical scores since a machine-learning model like Rebel Theorem only understands numbers in the end.</p><p>I won&#8217;t share the specific prompts we used because they&#8217;re both proprietary and quite technical, but the overall idea was to give the model clear instructions and as much relevant context as possible on the founders and their companies so it could score them accurately. Here&#8217;s the scoring scale we used for illustration:</p><blockquote><p>1&#8202;&#8212;&#8202;not related at all (e.g. founder worked in sales and product is a developer tool)</p></blockquote><blockquote><p>2&#8202;&#8212;&#8202;vaguely related (e.g. studied CS working and on a B2B vertical SaaS)</p></blockquote><blockquote><p>3&#8202;&#8212;&#8202;very related (e.g. worked as an AI researcher and is starting an AI start-up)</p></blockquote><blockquote><p>4&#8202;&#8212;&#8202;uniquely suited to be working on this startup (e.g. has a PHD or doctorate in biomedical devices and is creating a biomedical device start-up)</p></blockquote><p>Once we got the scores back from the AI model, we could start testing how predictive they are of YC startup outcomes.</p><p><strong>Step #3&#8202;&#8212; Test Scores</strong></p><p>Following our internal testing, the team prepared dozens of charts illustrating the usefulness of these new founder-product fit features.</p><p>Here are a few examples:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*hs2clNl0mb3IrTij4ssbDg@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*hs2clNl0mb3IrTij4ssbDg@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*hs2clNl0mb3IrTij4ssbDg@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*hs2clNl0mb3IrTij4ssbDg@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*hs2clNl0mb3IrTij4ssbDg@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*hs2clNl0mb3IrTij4ssbDg@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*hs2clNl0mb3IrTij4ssbDg@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*hs2clNl0mb3IrTij4ssbDg@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*hs2clNl0mb3IrTij4ssbDg@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*hs2clNl0mb3IrTij4ssbDg@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>This chart shows the frequency distribution of scores across each of the features that we tested&#8202;&#8212;&#8202;essentially how common different scores are for each feature. For example, we see it&#8217;s quite common for founders to have some background in the same geography of their startup&#8217;s target market, but it&#8217;s more rare for founders to have demonstrated adaptability moving between large-corporate environments and scrappy startup settings.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*Y5-xRgRltglPr82iRet6iA@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Y5-xRgRltglPr82iRet6iA@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Y5-xRgRltglPr82iRet6iA@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Y5-xRgRltglPr82iRet6iA@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*Y5-xRgRltglPr82iRet6iA@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*Y5-xRgRltglPr82iRet6iA@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*Y5-xRgRltglPr82iRet6iA@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Y5-xRgRltglPr82iRet6iA@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Y5-xRgRltglPr82iRet6iA@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Y5-xRgRltglPr82iRet6iA@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>This chart shows how predictive each founder-product fit feature is of startup success. The clear winners are education (i.e., how relevant the founder&#8217;s academic background to their startup) and leadership progression (how clearly the founder&#8217;s career path shows progression toward higher leadership responsibility), at least when the scores are very high or low, as scores in the middle are mostly noise.</p><p>Given the <a href="https://link.medium.com/D6Ootw8LdJb">steep power law</a> of venture capital returns and how rare great investments are, sometimes it&#8217;s less about picking winners than avoiding losers. So, we look at not only how good new features are at predicting startup success, but also at predicting which startups are most likely to fail.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*m7Nz8TpCcKckoOpO3b0yAg@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*m7Nz8TpCcKckoOpO3b0yAg@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*m7Nz8TpCcKckoOpO3b0yAg@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*m7Nz8TpCcKckoOpO3b0yAg@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*m7Nz8TpCcKckoOpO3b0yAg@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*m7Nz8TpCcKckoOpO3b0yAg@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*m7Nz8TpCcKckoOpO3b0yAg@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*m7Nz8TpCcKckoOpO3b0yAg@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*m7Nz8TpCcKckoOpO3b0yAg@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*m7Nz8TpCcKckoOpO3b0yAg@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>This chart shows how good each founder-product fit feature is at predicting which startups will likely fail&#8202;&#8212;&#8202;where they really seem to shine! Notice the higher dispersion in the whisker plots vs the previous success chart. A multivariate analysis found our founder-product fit scores are about 10x better at predicting which startups will likely fail than which will succeed.</p><p>What this tells us is a <em>lack</em> of founder-product fit is much more detrimental to a startup than an abundance is helpful. Companies that showed poor founder-product fit around geography, go-to market (i.e., how well the founder understands the startup&#8217;s go-to-market approach) and work experience (i.e., how relevant the founder&#8217;s prior work experience is to the company) were much more likely to die than their peers.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*syX0FyqgazvyA2XX8TDCyw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*syX0FyqgazvyA2XX8TDCyw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*syX0FyqgazvyA2XX8TDCyw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*syX0FyqgazvyA2XX8TDCyw.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*syX0FyqgazvyA2XX8TDCyw.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*syX0FyqgazvyA2XX8TDCyw.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*syX0FyqgazvyA2XX8TDCyw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*syX0FyqgazvyA2XX8TDCyw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*syX0FyqgazvyA2XX8TDCyw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*syX0FyqgazvyA2XX8TDCyw.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Finally, this chart shows the statistical significance of each feature when it comes to predicting startup failure, which as we&#8217;ve discussed, they&#8217;re better for than predicting success. In order for a feature to make it into our model, it needs to not only prove predictive of certain startup outcomes, but also with high statistical significance (i.e., beyond mere chance). When it comes to predicting startup failure, all of the founder-product fit features proved statistically significant, though to varying degrees.</p><p><strong>Step #4&#8202;&#8212; Retrain Model</strong></p><p>Once we&#8217;ve thoroughly tested new features and know them to be both powerful and statistically-significant predictors of YC startup outcomes, we retrain our core Rebel Theorem model incorporating these new features. This way the model can look at the new features alongside the 200+ other features it&#8217;s already been trained on to holistically predict YC startup outcomes, and we can see the full impact of the new features on overall model performance. The combined impact of these new founder-product fit scores proved quite dramatic, even though each feature taken in isolation had relatively little predictive power.</p><p>So, that&#8217;s how our new founder-product fit features made it into Rebel Theorem. When I think about how exhaustive and scientifically-rigorous our process is for adding new features to our model, and how incredibly predictive it is these days of YC startup outcomes, I wonder how I ever made investment decisions without it.</p>]]></content:encoded></item><item><title><![CDATA[On Rebel Theorem 4.0]]></title><description><![CDATA[Last year I shared a post On Rebel Theorem 3.0, which was the world&#8217;s most advanced machine-learning (ML) algorithm at the time for predicting Y Combinator startup success.]]></description><link>https://insights.rebelfund.vc/p/on-rebel-theorem-4-0-55d04b0732e3</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-rebel-theorem-4-0-55d04b0732e3</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Thu, 19 Jun 2025 18:22:06 GMT</pubDate><enclosure url="https://cdn-images-1.medium.com/max/800/1*Wc8ezsWlRdrElEGpU0yc2Q@2x.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>Last year I shared a post <a href="https://jaredheyman.medium.com/on-rebel-theorem-3-0-d33f5a5dad72">On Rebel Theorem 3.0</a>, which was the world&#8217;s most advanced machine-learning (ML) algorithm at the time for predicting <a href="https://www.ycombinator.com/">Y Combinator</a> startup success. Now a year later, <a href="https://www.rebelfund.vc">Rebel Fund</a> has released our latest Rebel Theorem 4.0 model, and I&#8217;m excited to share how the model works and performs, and what we learned developing it.</p><p>In case you&#8217;re new here, Rebel is one of the largest investors in the Y Combinator startup ecosystem, now with 250+ YC portfolio companies valued collectively in the tens of billions of dollars. As an extremely data-driven fund, we&#8217;ve built the world&#8217;s most comprehensive dataset on YC startups and founders outside of YC itself, now encompassing millions of data points across every YC company in history.</p><p>We&#8217;ve invested millions of dollars into collecting this data and training our internal ML (and now AI) algorithms, which give us a major edge in identifying which YC startups in each batch are most likely to become tomorrow&#8217;s unicorns. It&#8217;s also helped inspire the <a href="https://jaredheyman.medium.com">dozens of blog posts</a> I&#8217;ve published on YC startup trends, outcomes, and optimal investor strategies.</p><p><strong>Outcomes</strong></p><p>I&#8217;ll start by explaining the outcomes our new Rebel Theorem 4.0 model is designed to predict.</p><p>As I explained in my popular post <a href="https://link.medium.com/D6Ootw8LdJb">On the power law of Y Combinator startups</a>, the top ~6% of YC startups that become $1B+ companies drive the overwhelming majority of early investor returns. So, the obvious thing would be to predict startup valuation growth directly, but because the power law driving YC startup outcomes is so steep, the algorithm would only try to predict the extreme outliers (e.g. DoorDash, Airbnb, Stripe, Coinbase) which would add too much &#8216;noise&#8217; to be useful.</p><p>So instead, as with Rebel Theorem 3.0, we bucketed YC startups into three broad categories for model training purposes, which we <em>can</em> predict accurately and reliably:</p><p><strong>&#8220;Success&#8221;</strong>&#8202;&#8212;&#8202;$60M+ valuation and operating or exited</p><p><strong>&#8220;Zombie&#8221;</strong>&#8202;&#8212;&#8202;Under $60M valuation and still operating</p><p><strong>&#8220;Dead&#8221;</strong>&#8202;&#8212;&#8202;No longer operating nor exited</p><p>For batches before 2019, we&#8217;ve observed that 1 in 4 companies that achieved $60M+ valuations eventually became unicorns. So, if we can enrich our portfolio with more of these &#8216;Success&#8217; companies, we&#8217;ll end up with a disproportionate share of unicorns in the end, the true driver of portfolio returns.</p><p><strong>Performance</strong></p><p>Now let&#8217;s get to the exciting part&#8202;&#8212; how well our algorithm performs at predicting YC startup success.</p><p>The chart below shows the actual outcomes of startups with top-decile Rebel Theorem 4.0 scores in backtesting. As you&#8217;ll see, nearly 70% of startups predicted to be a Success actually <em>were</em> one&#8202;&#8212;&#8202;about <strong>2.5x better than YC averages</strong>. These startups were only half as likely to end up Dead as their peers, and only one-third as likely to end up Zombies.</p><p>This is performance is astonishing if you know how difficult it is to predict startup success at the seed stage. You&#8217;ll also see that our v4 algorithm performed much better than our year-old v3 algorithm, thanks to more and higher quality training data.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*Wc8ezsWlRdrElEGpU0yc2Q@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Wc8ezsWlRdrElEGpU0yc2Q@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Wc8ezsWlRdrElEGpU0yc2Q@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Wc8ezsWlRdrElEGpU0yc2Q@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*Wc8ezsWlRdrElEGpU0yc2Q@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*Wc8ezsWlRdrElEGpU0yc2Q@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*Wc8ezsWlRdrElEGpU0yc2Q@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Wc8ezsWlRdrElEGpU0yc2Q@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Wc8ezsWlRdrElEGpU0yc2Q@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Wc8ezsWlRdrElEGpU0yc2Q@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>If you&#8217;re an investor, you&#8217;re probably wondering how this outperformance vs YC averages might translate into financial returns. We backtested that as well:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*ErGWIr2xcTbK5jh-tNDLMQ@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*ErGWIr2xcTbK5jh-tNDLMQ@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*ErGWIr2xcTbK5jh-tNDLMQ@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*ErGWIr2xcTbK5jh-tNDLMQ@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*ErGWIr2xcTbK5jh-tNDLMQ@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*ErGWIr2xcTbK5jh-tNDLMQ@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*ErGWIr2xcTbK5jh-tNDLMQ@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*ErGWIr2xcTbK5jh-tNDLMQ@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*ErGWIr2xcTbK5jh-tNDLMQ@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*ErGWIr2xcTbK5jh-tNDLMQ@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>It&#8217;s a busy chart, here are the key findings:</p><ul><li><p><strong>Average YC Demo Day</strong> investor returns are estimated around a <strong>15&#8211;18% gross IRR</strong> for mature vintages&#8202;&#8212;&#8202;already better than an S&amp;P 500 index</p></li><li><p><strong>Y Combinator </strong>itself <em>kills it</em>&#8202;&#8212;&#8202;due to its super-low entry valuation, we estimate YC achieved around a <strong>45&#8211;55% gross IRR&#185;</strong> for mature vintages, placing it amongst the highest performing venture funds of all time</p></li><li><p><strong>Rebel Theorem 4.0</strong> <em>really kills it</em>&#8202;&#8212;&#8202;had we invested in the top 10% of YC startups per the algorithm, our portfolio would have achieved an estimated <strong>65%+ gross IRR</strong> for these mature vintages, even better than YC itself</p></li></ul><p>I know some of these IRR numbers sound fanciful, but in an asset class where a single big winner can return 1,000x or more, they&#8217;re entirely possible with great deal selection, access and portfolio strategy.</p><p><strong>Top Features</strong></p><p>One of the questions I&#8217;m asked most often is which characteristics (or &#8216;features&#8217; in data science speak) are most predictive of YC startup success.</p><p>At Rebel, we looked at hundreds of them when training our models, and our latest Rebel Theorem 4.0 model factors 200+ features into its scores. We scoured the internet and various public and private databases for <em>any</em> quantifiable characteristic of YC startups and founders that may have predictive power, ranging from the obvious (previous founder successes, educational and work history, company location, etc) to the obscure (whether founders have last names associated with certain nationalities, if they wear glasses in their LinkedIn profile picture, who is their YC Group Partner, etc)</p><p>Here are the top 25 features that our model found most predictive of YC startup success:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*o0JUE0AC_xCdSquhHipzhw@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*o0JUE0AC_xCdSquhHipzhw@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*o0JUE0AC_xCdSquhHipzhw@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*o0JUE0AC_xCdSquhHipzhw@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*o0JUE0AC_xCdSquhHipzhw@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*o0JUE0AC_xCdSquhHipzhw@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*o0JUE0AC_xCdSquhHipzhw@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*o0JUE0AC_xCdSquhHipzhw@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*o0JUE0AC_xCdSquhHipzhw@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*o0JUE0AC_xCdSquhHipzhw@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>To make it more readable, we grouped these top features into 4 color-coded categories:</p><ul><li><p><em>Company</em>&#8202;&#8212; features related to the company itself (company location, age, sector, etc)</p></li><li><p><em>Education</em>&#8202;&#8212; features related to the founders&#8217; educational history (where they went to school, how long they studied, how long ago they graduated, etc)</p></li><li><p><em>Employment</em>&#8202;&#8212; features related to the founders&#8217; employment history (years of work experience, years of co-founder experience, cities worked in, etc)</p></li><li><p><em>Personality</em>&#8202;&#8212; features related to the founders&#8217; personality traits, as inferred by a specialized AI model based on their online footprint (skepticism, dominance, risk aversion, etc)</p></li></ul><p>You&#8217;ll notice the vast majority of these top features are related to the <em>founders</em> vs the company itself. This is in part because we simply know more about the founders than the company at the seed stage, and in part because founders really are the main driver of YC startup success. I often say that technically our algorithm is scoring companies, but really it&#8217;s scoring founders.</p><p>You also may notice that many of the most predictive features are obvious&#8202;&#8212;&#8202;things like how experienced the founders are (both as co-founders and overall), how old they are (years since they started college or their first job are proxies), or where they went to school (top school rank).</p><p>However, there is a long tail of highly predictive features that are less obvious, like certain personality traits, and even how long the founders have lived in San Francisco.</p><p>I should note that since this is a non-linear model, you shouldn&#8217;t assume that &#8216;more is better&#8217; of any given feature. The top 25 list above is based on <em>feature frequency</em> in the 100+ decision trees our model uses to score companies, not how &#8216;good&#8217; or &#8216;bad&#8217; a given feature is.</p><p>For example, while years of founder work experience is the #1 most predictive feature, this could mean that more years are generally better, more years are generally worse, or more likely, that it depends on various other features in the decision tree (e.g., more years of work experience is better for healthcare founders, worse for AI founders, better for fintech founders who have achieved an exit in the past, etc)</p><p>Non-linear ML models like Rebel Theorem 4.0 are unfathomably complex and difficult to explain, but essentially the model has figured out which of the hundreds of features in its training data are most predictive of YC startup success, in which combinations, for which types or companies. For example, if it&#8217;s statistically true that younger, non-skeptical, highly-conscientious, top university educated founders, who worked at certain prior employers, and co-founded a certain number of companies in the past, tend to build the most successful B2B startups, but not B2C startups, the algorithm would &#8216;know&#8217; that and score them accordingly.</p><p>This is the reason that a sophisticated ML model like Rebel Theorem can achieve investment returns far surpassing any mortal, and why we&#8217;ve invested so heavily into its underlying data collection, training and development&#8202;&#8212;&#8202;the algorithm literally has super-human capabilities&#178;.</p><p><strong>Other Features</strong></p><p>We want to give our algorithm the best chance possible of discovering non-obvious yet statistically-valid predictors of YC startup success, so we&#8217;ve been creative with the types of data we&#8217;ve included in its training set. Now that we&#8217;re armed with sophisticated AI platforms with computer vision, web search, and advanced reasoning, we can be more creative than ever.</p><p>I won&#8217;t divulge the full list of features Rebel Theorem 4.0 is trained on, but one of my favorite new additions is a series of features based on YC founders&#8217; LinkedIn profile pictures. Here are few fun examples:</p><p><em>Is the person&#8217;s hair neatly and professionally styled?</em></p><p>True examples:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*bmuBTdqhjXgpBse-8dRwUQ@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*bmuBTdqhjXgpBse-8dRwUQ@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*bmuBTdqhjXgpBse-8dRwUQ@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*bmuBTdqhjXgpBse-8dRwUQ@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*bmuBTdqhjXgpBse-8dRwUQ@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*bmuBTdqhjXgpBse-8dRwUQ@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*bmuBTdqhjXgpBse-8dRwUQ@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*bmuBTdqhjXgpBse-8dRwUQ@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*bmuBTdqhjXgpBse-8dRwUQ@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*bmuBTdqhjXgpBse-8dRwUQ@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>False examples:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*Ehx1LKk9MnkNtd6rRldyVQ@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Ehx1LKk9MnkNtd6rRldyVQ@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Ehx1LKk9MnkNtd6rRldyVQ@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Ehx1LKk9MnkNtd6rRldyVQ@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*Ehx1LKk9MnkNtd6rRldyVQ@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*Ehx1LKk9MnkNtd6rRldyVQ@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*Ehx1LKk9MnkNtd6rRldyVQ@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Ehx1LKk9MnkNtd6rRldyVQ@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Ehx1LKk9MnkNtd6rRldyVQ@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Ehx1LKk9MnkNtd6rRldyVQ@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p><em>Does this image appear to be a staged professional photo rather than a candid photo?</em></p><p>True examples:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*Cu6WQFMdT8oBjcZgxUsf3Q@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Cu6WQFMdT8oBjcZgxUsf3Q@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Cu6WQFMdT8oBjcZgxUsf3Q@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Cu6WQFMdT8oBjcZgxUsf3Q@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*Cu6WQFMdT8oBjcZgxUsf3Q@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*Cu6WQFMdT8oBjcZgxUsf3Q@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*Cu6WQFMdT8oBjcZgxUsf3Q@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Cu6WQFMdT8oBjcZgxUsf3Q@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Cu6WQFMdT8oBjcZgxUsf3Q@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Cu6WQFMdT8oBjcZgxUsf3Q@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>False example:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*_0K4NatfDuiVGocki7XmUg@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*_0K4NatfDuiVGocki7XmUg@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*_0K4NatfDuiVGocki7XmUg@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*_0K4NatfDuiVGocki7XmUg@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*_0K4NatfDuiVGocki7XmUg@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*_0K4NatfDuiVGocki7XmUg@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*_0K4NatfDuiVGocki7XmUg@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*_0K4NatfDuiVGocki7XmUg@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*_0K4NatfDuiVGocki7XmUg@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*_0K4NatfDuiVGocki7XmUg@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p><em>Does this person have a clean-shaven appearance or facial hair?</em></p><p>True example:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*N6l9QEr8_aV5OE3cZKaK5g@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*N6l9QEr8_aV5OE3cZKaK5g@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*N6l9QEr8_aV5OE3cZKaK5g@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*N6l9QEr8_aV5OE3cZKaK5g@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*N6l9QEr8_aV5OE3cZKaK5g@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*N6l9QEr8_aV5OE3cZKaK5g@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*N6l9QEr8_aV5OE3cZKaK5g@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*N6l9QEr8_aV5OE3cZKaK5g@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*N6l9QEr8_aV5OE3cZKaK5g@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*N6l9QEr8_aV5OE3cZKaK5g@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>False example:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*68HI0PZ3ylQQX7Ojcc2DOQ@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*68HI0PZ3ylQQX7Ojcc2DOQ@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*68HI0PZ3ylQQX7Ojcc2DOQ@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*68HI0PZ3ylQQX7Ojcc2DOQ@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*68HI0PZ3ylQQX7Ojcc2DOQ@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*68HI0PZ3ylQQX7Ojcc2DOQ@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*68HI0PZ3ylQQX7Ojcc2DOQ@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*68HI0PZ3ylQQX7Ojcc2DOQ@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*68HI0PZ3ylQQX7Ojcc2DOQ@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*68HI0PZ3ylQQX7Ojcc2DOQ@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Believe it not, some of these LinkedIn profile pic features are quite statistically predictive!</p><p><strong>Our AI Future</strong></p><p>We&#8217;ve started to incorporate even more sophisticated features into our model as well, which were only unlocked once advanced AI reasoning models were released like OpenAI&#8217;s new o3 model. These features allow us to quantify more subjective aspects of a startup, such as its product quality, the timeliness of its idea, competitive defensibility, founder-product fit, etc. that were until now only within the grasp of human experts.</p><p>In my post <a href="https://jaredheyman.medium.com/on-why-ai-is-coming-for-my-job-next-bb628166259c">On why AI is coming for my job next</a>, I shared the chart below illustrating how OpenAI&#8217;s o1 model (the predecessor to o3) made better predictions about the success of Rebel portfolio companies than our own internal partner ratings:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*LCWhLXIOmGxxTSoU063VZw@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*LCWhLXIOmGxxTSoU063VZw@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*LCWhLXIOmGxxTSoU063VZw@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*LCWhLXIOmGxxTSoU063VZw@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*LCWhLXIOmGxxTSoU063VZw@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*LCWhLXIOmGxxTSoU063VZw@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*LCWhLXIOmGxxTSoU063VZw@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*LCWhLXIOmGxxTSoU063VZw@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*LCWhLXIOmGxxTSoU063VZw@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*LCWhLXIOmGxxTSoU063VZw@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>I published that post a couple of months ago and it&#8217;s already outdated&#8202;&#8212;&#8202;the new o3 model does even better, and when o4 or GPT-5 is released, I&#8217;m sure it will perform better still.</p><p>Our &#8216;traditional&#8217; ML approach to predicting YC startup success has given us a massive advantage over other investors, and I expect our new advanced AI reasoning features to expand our advantage further. I&#8217;m convinced that our latest ML/AI models have already surpassed the capabilities of not only the average Silicon Valley VC, but <em>the very best</em> VCs&#8202;&#8212;&#8202;and as Sam Altman famously says &#8220;this is the dumbest these models will ever be&#8221;.</p><p>&#185;We estimate YC fund performance based on publicly announced and algorithmically estimated valuations of YC startups across these vintages. We have no special insight into actual YC fund performance.</p><p>&#178;As a thought experiment, for a human investor to effectively compete with with a non-linear ML model like Rebel Theorem, she would have to read millions of unique facts on thousands of YC startups and founders, memorize them, hypothesize and test hundreds of fact patterns that predict which startups are most likely to succeed, calculate their statistical significance, validate them in backtesting, memorize those patterns, and apply them to new startups and founders she sees to accurately calculate their odds of success&#8230; in a matter of seconds. Good luck!</p>]]></content:encoded></item><item><title><![CDATA[On YC startup exits (2025 update)]]></title><description><![CDATA[One of the questions I&#8217;m asked most often by Rebel Fund LPs is &#8220;How long will it take to get my money back?&#8221;]]></description><link>https://insights.rebelfund.vc/p/on-yc-startup-exits-2025-update-c6017e8e526e</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-yc-startup-exits-2025-update-c6017e8e526e</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Thu, 29 May 2025 22:40:36 GMT</pubDate><enclosure url="https://cdn-images-1.medium.com/max/800/1*YUjjC4lIlViow2u65lVm4g.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>One of the questions I&#8217;m asked most often by <a href="https://www.rebelfund.vc/">Rebel Fund</a> LPs is &#8220;How long will it take to get my money back?&#8221;</p><p>I get it&#8202;&#8212;&#8202;venture is an illiquid asset class with long holding periods, and especially with the recent slowdown in IPO and acquisition activity, DPI is top-of-mind for many venture investors.</p><p>A couple of years ago, I wrote a blog post <a href="https://jaredheyman.medium.com/on-yc-startup-exits-6d6ec04b14d9">On YC startup exits</a> that attempted to answer that question statistically. Rebel has improved our data infrastructure dramatically since then, allowing us to answer this with even more nuance, so I figured now is a good time for an update.</p><p>As my long-time readers know, Rebel maintains the most comprehensive dataset that exists of YC startups outside of YC itself, now with millions of datapoints across every YC founder and company in history, including the outcomes of every exited company (exit type, exit year, estimated valuation, etc). We use this data primarily to train our proprietary <a href="https://link.medium.com/lPJ0A1Zc4Kb">Rebel Theorem 3.0</a> machine learning algorithm, but it&#8217;s also useful for tracking interesting trends in the YC startup ecosystem.</p><p>So, I queried our database for every exited YC startup to date, excluding those with very small outcomes for early investors (&lt;10% IRR) to remove the acqui-hires, leaving me with ~250 &#8216;real&#8217; exits to analyze. As I mentioned in my popular post <a href="https://jaredheyman.medium.com/on-the-life-and-death-of-y-combinator-startups-d58aa03421f0">On the life and death of Y Combinator startups</a>, about ~40% of YC startups ultimately achieve an exit, so these ~250 companies are a small fraction of the thousands of YC startups that will eventually be acquired or go public.</p><p>The first thing I found is the median time it takes for a YC startup to achieve an exit is about 4 years, but not all exits are created equal. As you&#8217;ll see in the chart below, the &lt;$100M exits take only about 3 years, the $100M-$999M exits take closer to 6 years, and the $1B+ exits take nearly a decade!</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*YUjjC4lIlViow2u65lVm4g.png 424w, https://cdn-images-1.medium.com/max/800/1*YUjjC4lIlViow2u65lVm4g.png 848w, https://cdn-images-1.medium.com/max/800/1*YUjjC4lIlViow2u65lVm4g.png 1272w, https://cdn-images-1.medium.com/max/800/1*YUjjC4lIlViow2u65lVm4g.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*YUjjC4lIlViow2u65lVm4g.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*YUjjC4lIlViow2u65lVm4g.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*YUjjC4lIlViow2u65lVm4g.png 424w, https://cdn-images-1.medium.com/max/800/1*YUjjC4lIlViow2u65lVm4g.png 848w, https://cdn-images-1.medium.com/max/800/1*YUjjC4lIlViow2u65lVm4g.png 1272w, https://cdn-images-1.medium.com/max/800/1*YUjjC4lIlViow2u65lVm4g.png 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p>The reason the overall median time to exit is just 4 years is the smaller exits are much more common:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*wwmCdyVHp8V1uqrDgCK4Fg.png 424w, https://cdn-images-1.medium.com/max/800/1*wwmCdyVHp8V1uqrDgCK4Fg.png 848w, https://cdn-images-1.medium.com/max/800/1*wwmCdyVHp8V1uqrDgCK4Fg.png 1272w, https://cdn-images-1.medium.com/max/800/1*wwmCdyVHp8V1uqrDgCK4Fg.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*wwmCdyVHp8V1uqrDgCK4Fg.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*wwmCdyVHp8V1uqrDgCK4Fg.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*wwmCdyVHp8V1uqrDgCK4Fg.png 424w, https://cdn-images-1.medium.com/max/800/1*wwmCdyVHp8V1uqrDgCK4Fg.png 848w, https://cdn-images-1.medium.com/max/800/1*wwmCdyVHp8V1uqrDgCK4Fg.png 1272w, https://cdn-images-1.medium.com/max/800/1*wwmCdyVHp8V1uqrDgCK4Fg.png 1456w" sizes="100vw"></picture><div></div></div></div></figure></div><p>In my post <a href="https://link.medium.com/D6Ootw8LdJb">On the power law of Y Combinator startups</a>, I spoke about how incredibly skewed startup investor returns are towards the big winners, and we see this when looking at exited startups as well. Even though the unicorns are just ~8% of all exits, they represent a whopping ~93% of the cash returned to YC startup seed investors:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*gJfygpqdne_58VsGLZMc7Q.png 424w, https://cdn-images-1.medium.com/max/800/1*gJfygpqdne_58VsGLZMc7Q.png 848w, https://cdn-images-1.medium.com/max/800/1*gJfygpqdne_58VsGLZMc7Q.png 1272w, https://cdn-images-1.medium.com/max/800/1*gJfygpqdne_58VsGLZMc7Q.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*gJfygpqdne_58VsGLZMc7Q.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*gJfygpqdne_58VsGLZMc7Q.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*gJfygpqdne_58VsGLZMc7Q.png 424w, https://cdn-images-1.medium.com/max/800/1*gJfygpqdne_58VsGLZMc7Q.png 848w, https://cdn-images-1.medium.com/max/800/1*gJfygpqdne_58VsGLZMc7Q.png 1272w, https://cdn-images-1.medium.com/max/800/1*gJfygpqdne_58VsGLZMc7Q.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Since exit value is so skewed towards the unicorns, even though the median startup exit takes only 4 years, the median <em><strong>exit dollar</strong></em> takes closer to 9 years. Our data shows that ~90 % of exits occur in &#8804; 9 years, but those account for only ~50 % of the exit dollars. The last ~10 % of exits&#8202;&#8212;&#8202;mostly unicorn-scale IPOs&#8202;&#8212;&#8202;deliver the other half of the money, and they bring up the median exit dollar timeline.</p><p>Predictably, the IRR enjoyed by seed investors in companies achieving these larger exits is much higher, hitting 75% IRR for the unicorns, which is of course insanely good:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*M2TrdUYl8n5GoMc0UBRzKw.png 424w, https://cdn-images-1.medium.com/max/800/1*M2TrdUYl8n5GoMc0UBRzKw.png 848w, https://cdn-images-1.medium.com/max/800/1*M2TrdUYl8n5GoMc0UBRzKw.png 1272w, https://cdn-images-1.medium.com/max/800/1*M2TrdUYl8n5GoMc0UBRzKw.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*M2TrdUYl8n5GoMc0UBRzKw.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*M2TrdUYl8n5GoMc0UBRzKw.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*M2TrdUYl8n5GoMc0UBRzKw.png 424w, https://cdn-images-1.medium.com/max/800/1*M2TrdUYl8n5GoMc0UBRzKw.png 848w, https://cdn-images-1.medium.com/max/800/1*M2TrdUYl8n5GoMc0UBRzKw.png 1272w, https://cdn-images-1.medium.com/max/800/1*M2TrdUYl8n5GoMc0UBRzKw.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Many investors have felt the pace of startup exits has slowed recently, and we see this reflected in our data as well. After the craziness of the 2021 ZIRP-era peak, YC exits have been slower in recent years than as far back as 2017, despite steadily growing batch sizes:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*ZHYcinrkWh1JT0POUm7KVg.png 424w, https://cdn-images-1.medium.com/max/800/1*ZHYcinrkWh1JT0POUm7KVg.png 848w, https://cdn-images-1.medium.com/max/800/1*ZHYcinrkWh1JT0POUm7KVg.png 1272w, https://cdn-images-1.medium.com/max/800/1*ZHYcinrkWh1JT0POUm7KVg.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*ZHYcinrkWh1JT0POUm7KVg.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*ZHYcinrkWh1JT0POUm7KVg.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*ZHYcinrkWh1JT0POUm7KVg.png 424w, https://cdn-images-1.medium.com/max/800/1*ZHYcinrkWh1JT0POUm7KVg.png 848w, https://cdn-images-1.medium.com/max/800/1*ZHYcinrkWh1JT0POUm7KVg.png 1272w, https://cdn-images-1.medium.com/max/800/1*ZHYcinrkWh1JT0POUm7KVg.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>We also found that median exit times vary quite a bit by technology sector, with Industrials, Consumer, and B2B companies achieving quicker exits, and Healthcare companies taking the longest:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*nIzEoyGTUaZ6u2f-qYDRqw.png 424w, https://cdn-images-1.medium.com/max/800/1*nIzEoyGTUaZ6u2f-qYDRqw.png 848w, https://cdn-images-1.medium.com/max/800/1*nIzEoyGTUaZ6u2f-qYDRqw.png 1272w, https://cdn-images-1.medium.com/max/800/1*nIzEoyGTUaZ6u2f-qYDRqw.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*nIzEoyGTUaZ6u2f-qYDRqw.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*nIzEoyGTUaZ6u2f-qYDRqw.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*nIzEoyGTUaZ6u2f-qYDRqw.png 424w, https://cdn-images-1.medium.com/max/800/1*nIzEoyGTUaZ6u2f-qYDRqw.png 848w, https://cdn-images-1.medium.com/max/800/1*nIzEoyGTUaZ6u2f-qYDRqw.png 1272w, https://cdn-images-1.medium.com/max/800/1*nIzEoyGTUaZ6u2f-qYDRqw.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>So what does all this mean for early-stage startup investors? Here&#8217;s my take:</p><ul><li><p><strong>Power-law is real</strong>&#8202;&#8212;&#8202;Investors looking for DPI should follow the same strategies as those seeking IRR, which I speak to in my <a href="https://link.medium.com/D6Ootw8LdJb">power law post</a></p></li><li><p><strong>Be patient, my friends</strong>&#8202;&#8212;&#8202;The best exits take the longest and drive the vast majority of cash returns, so ride your winners all the way</p></li><li><p><strong>Keep a balanced diet&#8202;</strong>&#8212;&#8202;Invest in a wide variety of sectors to cover all your bases. B2B companies exit more quickly and often, Fintech companies post the highest IRR, Consumer companies drive the most exit dollars but are riskier, etc.</p></li><li><p><strong>Don&#8217;t try to time the market</strong>&#8202;&#8212;&#8202;Nobody knows what the IPO market will look like a decade from now when today&#8217;s best seed investments are exiting, nor when the next 2021-style liquidity bonanza will hit, so just keep making startup investments year after year</p></li><li><p><strong>Don&#8217;t discount unrealized IRR&#8202;</strong>&#8212;&#8202;I know you can&#8217;t buy beer or yachts with unrealized returns, but they definitely still &#8216;count&#8217; in an asset class like venture with very long liquidity timelines, and are useful for tracking portfolio performance along the way</p></li></ul><p>As always, I hope you guys found this helpful. Best of luck investing in tomorrow&#8217;s 10+ year overnight successes!</p><p><strong>Update (07/10/2025)</strong></p><p>This post got very popular, and I&#8217;m often asked by readers what our data shows actual realized DPI looks like historically for YC Demo Day investors X years after any given seed investment. So, to answer that question, I queried our database to look at only YC startups from the 2008&#8211;2013 batch years, so they&#8217;re old enough to have mostly exited, yet still reflect relatively recent YC history. I came up with the chart below:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*MJyblpa-vcbydvQnvJPtzw@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*MJyblpa-vcbydvQnvJPtzw@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*MJyblpa-vcbydvQnvJPtzw@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*MJyblpa-vcbydvQnvJPtzw@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*MJyblpa-vcbydvQnvJPtzw@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*MJyblpa-vcbydvQnvJPtzw@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*MJyblpa-vcbydvQnvJPtzw@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*MJyblpa-vcbydvQnvJPtzw@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*MJyblpa-vcbydvQnvJPtzw@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*MJyblpa-vcbydvQnvJPtzw@2x.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>This leads us the following conclusions:</p><p>&#8212; Even though unrealized returns grow steadily as companies raise growth rounds at higher valuations, DPI takes a <em>long</em> time to materialize, which makes sense given the best exits take the longest.</p><p>&#8212; YC Demo Day investors shouldn&#8217;t expect to get their money back (1x DPI) around 6 years after any given investment cohort, unless of course they sell early on the secondary market. However, cash returns start to grow exponentially from there as the big winners ($1B+ companies) start to achieve liquidity.</p><p>&#8212; Realized returns get really exciting around Year 10 (~20x DPI net of dilution) but these returns are <em>extremely</em> outlier-driven. To illustrate, I added a dotted blue line excluding the $10B+ companies, which brings the Year 10 DPI to a much more modest ~2.5x</p><p>&#8212; Even though I cut off this chart at Year 10, in reality, DPI continues to grow exponentially for another couple of years if the $10B+ companies remain included.</p><p>This is another illustration of why all roads point to a large portfolio strategy when it comes to YC startup seed investing. When the <a href="https://link.medium.com/D6Ootw8LdJb">power law</a> is this steep, you simply can&#8217;t afford to miss the mega-hits.</p>]]></content:encoded></item><item><title><![CDATA[On investing in “hot” Y Combinator startups]]></title><description><![CDATA[I was having a debate recently with a fellow Y Combinator startup investor on whether or not companies that get &#8220;hot&#8221; at Demo Day make for good seed investments.]]></description><link>https://insights.rebelfund.vc/p/on-investing-in-hot-y-combinator-startups-aed26869f77c</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-investing-in-hot-y-combinator-startups-aed26869f77c</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Fri, 04 Apr 2025 20:56:34 GMT</pubDate><enclosure url="https://cdn-images-1.medium.com/max/800/1*FxJcOlBo9cARi7iPlermIQ.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>I was having a debate recently with a fellow Y Combinator startup investor on whether or not companies that get &#8220;hot&#8221; at Demo Day make for good seed investments. These are the companies that, for whatever reason, get very popular with investors and find their seed rounds heavily oversubscribed. The founders of these companies become rationally quite selective of their investors, and irrationally confident in their startup&#8217;s future despite its nascent stage.</p><p>Anecdotally, several YC partners have told me they haven&#8217;t noticed a strong relationship between companies that get &#8220;hot&#8221; at Demo Day and those that ultimately have good outcomes. There are plenty of examples of companies that were hot during their YC batch and did quite well (Stripe, Gusto, etc), many examples of companies that were <em>not</em> popular with investors and did quite well (Segment, Deel, etc), and of course, many companies that were hot at Demo Day and did very poorly (I won&#8217;t name names)</p><p>Since <a href="https://www.rebelfund.vc/%5C">Rebel Fund</a> has developed the largest and most comprehensive dataset that exists of Y Combinator startups, founders, and outcomes outside of YC itself, it only seemed fitting that we should answer this question statistically, and this blog post will explain what we found.</p><p>To start, we randomly sampled a couple hundred YC startups from the 2015&#8211;2018 batch years, roughly evenly divided across each of the following categories based on their latest valuations:</p><p>Unicorns&#8202;&#8212;&#8202;$1B+ valuations</p><p>Top companies&#8202;&#8212;&#8202;$100M+ valuations</p><p>Regular startups&#8202;&#8212;&#8202;&lt;$100M valuations or inactive</p><p>We choose the 2015&#8211;2018 batch years since these companies are old enough to have matured, yet young enough to reflect relatively recent YC history. We chose companies spread across these valuation categories so we&#8217;d have a good dispersement of outcomes within our sample.</p><p>Since we can&#8217;t measure a company&#8217;s &#8220;hotness&#8221; at Demo Day directly, we used as a proxy the total amount of capital each company raised in their batch year&#8202;&#8212;&#8202;one of the hundreds of metrics Rebel tracks on each YC startup. Our rationale is the hotter a company was at Demo Day, the more capital they probably raised.</p><p>The first thing we found is there is indeed a decent positive correlation between the total amount of capital a company raised in their batch year and their latest valuation (0.23):</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*FxJcOlBo9cARi7iPlermIQ.png 424w, https://cdn-images-1.medium.com/max/800/1*FxJcOlBo9cARi7iPlermIQ.png 848w, https://cdn-images-1.medium.com/max/800/1*FxJcOlBo9cARi7iPlermIQ.png 1272w, https://cdn-images-1.medium.com/max/800/1*FxJcOlBo9cARi7iPlermIQ.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*FxJcOlBo9cARi7iPlermIQ.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*FxJcOlBo9cARi7iPlermIQ.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*FxJcOlBo9cARi7iPlermIQ.png 424w, https://cdn-images-1.medium.com/max/800/1*FxJcOlBo9cARi7iPlermIQ.png 848w, https://cdn-images-1.medium.com/max/800/1*FxJcOlBo9cARi7iPlermIQ.png 1272w, https://cdn-images-1.medium.com/max/800/1*FxJcOlBo9cARi7iPlermIQ.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>The correlation wasn&#8217;t nearly as strong as that between their total funding raised and latest valuation (0.64) but it&#8217;s still decent. So, at first blush, hotter companies do indeed perform a bit better.</p><p>We then dug in a bit deeper, looking at the relationship on a logarithmic scale (since YC startup outcomes tend to follow a <a href="https://link.medium.com/D6Ootw8LdJb">power law</a>) and while you can still see the positive relationship between total capital raised in batch year vs latest company valuation, the slope elasticity is only 1.19, meaning that for every 1% increase in batch-year capital raised, there is only a 1.19% increase latest company valuation:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*Mk8bDydSw49lPV1J39T98A.png 424w, https://cdn-images-1.medium.com/max/800/1*Mk8bDydSw49lPV1J39T98A.png 848w, https://cdn-images-1.medium.com/max/800/1*Mk8bDydSw49lPV1J39T98A.png 1272w, https://cdn-images-1.medium.com/max/800/1*Mk8bDydSw49lPV1J39T98A.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*Mk8bDydSw49lPV1J39T98A.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*Mk8bDydSw49lPV1J39T98A.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*Mk8bDydSw49lPV1J39T98A.png 424w, https://cdn-images-1.medium.com/max/800/1*Mk8bDydSw49lPV1J39T98A.png 848w, https://cdn-images-1.medium.com/max/800/1*Mk8bDydSw49lPV1J39T98A.png 1272w, https://cdn-images-1.medium.com/max/800/1*Mk8bDydSw49lPV1J39T98A.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>YC startups tend to increase their valuations when investor demand is high, targeting a specific percentage of seed round dilution, so companies that raised more in their batch year were likely proportionately more expensive to investors. So, after taking into account their higher entry valuations, an investor would only expect a ~19% bump in ROI with a &#8220;hot&#8221; company.</p><p>Also, the R&#178; is only 0.090, which means only ~9% of the variation in a company&#8217;s latest valuation is explained by their batch-year funding, so the other ~91% is explained by other factors.</p><p>The bottom line is that a company being &#8220;hot&#8221; with investors at Demo Day is generally a positive signal, but a relatively weak one, especially once you take into account their higher entry valuation. So, the implication for investors is don&#8217;t be afraid to invest in hot companies when you like them, since they do perform slightly better than their peers, but don&#8217;t assume they&#8217;ll do well just because they&#8217;re hot, since &gt;90% of a company&#8217;s success is driven by other factors.</p><p>This is why at Rebel, we mostly ignore deal heat and instead focus on the fundamentals of each company, like its team, product, market, defensibility, etc that predict the other &gt;90% of a company&#8217;s success. As I explained my last post <a href="https://jaredheyman.medium.com/on-why-ai-is-coming-for-my-job-next-bb628166259c">On why AI is coming for my job next</a>, we&#8217;d rather use our vast database, cutting-edge <a href="https://link.medium.com/lPJ0A1Zc4Kb">Rebel Theorem algorithm</a>, and experienced partners to select tomorrow&#8217;s YC unicorns&#8202;&#8212;&#8202;not just follow the herd.</p>]]></content:encoded></item><item><title><![CDATA[On why AI is coming for my job next]]></title><description><![CDATA[On why AI is coming for my job next]]></description><link>https://insights.rebelfund.vc/p/on-why-ai-is-coming-for-my-job-next-bb628166259c</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-why-ai-is-coming-for-my-job-next-bb628166259c</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Tue, 01 Apr 2025 21:48:45 GMT</pubDate><enclosure url="https://cdn-images-1.medium.com/max/800/1*1HKIrS1Om6DsPu6nWuQw1A@2x.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<h2>On why AI is coming for my job&nbsp;next</h2><p>I&#8217;ve written <a href="https://jaredheyman.medium.com/">dozens of posts</a> over the years on Y Combinator startup trends and how we use data and algorithms at <a href="https://www.rebelfund.vc">Rebel Fund</a> to invest in the top 10% of new YC companies. We&#8217;ve invested millions of dollars into our data automation infrastructure, proprietary machine learning algorithms, and internal software to quickly identify and invest in the best new YC startups each year. Not only do we invest in hundreds of the most cutting-edge technology startups coming out of Silicon Valley, but we also invest in developing and using the most cutting-edge technology possible to do so.</p><p>It should thus be no surprise that we&#8217;ve been keeping a close eye on the latest developments in AI, and figuring out how to merge its fast growing capabilities into our existing technology infrastructure. In this post, I&#8217;ll explain some of the mind-blowing capabilities that AI has developed over just the past few months for investors, and why I think in the near future, early-stage investors who don&#8217;t leverage AI in their due diligence and selection process will be left in the dust by those who do.</p><p>First off, if you&#8217;re not already familiar with our proprietary <a href="https://jaredheyman.medium.com/on-rebel-theorem-3-0-d33f5a5dad72">Rebel Theorem 3.0</a> machine learning algorithm, I&#8217;d suggest reading that post first, since it gives some background on how we&#8217;ve used millions of data points to train an ML model to accurately predict seed-stage YC startup success.</p><p>When we developed the previous version <a href="https://jaredheyman.medium.com/on-rebel-theorem-2-0-51404e7a6cde">Rebel Theorem 2.0</a> in 2022, we were limited to using data features that were either easily quantified (like years of founder education or work experience) or that could be easily coded numerically (like where a founder studied or worked in the past). By the time we released Rebel Theorem 3.0 in 2024, we could use &#8216;older&#8217; AI models like GPT-4 to incorporate new AI-generated data features, like inferred founder personality profiles or demographic characteristics based on other data we had collected on them. We found that each time we re-trained the algorithm with deeper data insights on companies and founders, the more accurately it could predict startup outcomes.</p><p>The most recent advance in AI is a new series of reasoning models, like OpenAI&#8217;s o1 and o3 models, which can think more deeply, logically, and flexibly than their predecessors. These have unleashed yet another set of new features that we can incorporate into our Rebel Theorem algorithm, bringing its predictive performance (and presumable the quality of investments we can make with it) to even greater heights.</p><p>These models have created an entirely new set of capabilities for early-stage startup investors who know how to use them properly. Because they take the time to think through inputs logically and consistently, given the right training, prompts and context, they can accurately answer more nuanced and complex questions like &#8220;Does the CAC to LTV ratio for this startup make sense given its GTM strategy?&#8221; or &#8220;Is this founding team&#8217;s experience aligned with their product vision?&#8221;</p><p>In our early testing of new AI reasoning models&#8217; output, we&#8217;ve been blown away by not only their performance vs previous generation models, but also our own human experts (including me). Below is a humbling chart that slows the correlation between OpenAI&#8217;s o1 reasoning model&#8217;s predictions of Rebel portfolio startups valuation growth vs our internal partner ratings, based solely on information available on each startup at the time we invested, presented to the model in a certain structured format that we developed internally:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*1HKIrS1Om6DsPu6nWuQw1A@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*1HKIrS1Om6DsPu6nWuQw1A@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*1HKIrS1Om6DsPu6nWuQw1A@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*1HKIrS1Om6DsPu6nWuQw1A@2x.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*1HKIrS1Om6DsPu6nWuQw1A@2x.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*1HKIrS1Om6DsPu6nWuQw1A@2x.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*1HKIrS1Om6DsPu6nWuQw1A@2x.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*1HKIrS1Om6DsPu6nWuQw1A@2x.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*1HKIrS1Om6DsPu6nWuQw1A@2x.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*1HKIrS1Om6DsPu6nWuQw1A@2x.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p>At Rebel, we&#8217;re a top-tier venture fund with some of the smartest partners in Silicon Valley, co-founding companies now valued over $100B in total and investing of hundreds of startups together, yet a properly trained AI reasoning model fed the proper context on a company and its founders, with the proper format and prompting, can make better predictions of startup outcomes than even us. In fact, I don&#8217;t think <em>any</em> human investor can make such highly accurate valuation growth predictions at such an early stage of a startup&#8217;s lifecycle.</p><p>Even more impressive than the model consistently outperforming ratings from me and my partners is that its predictions show an increasingly strong correlation with startup valuation growth for earlier vintages, which have had more time to mature (or, in data science terms, more time for valuation outcomes to disperse). This aligns perfectly with what we&#8217;d expect from a model genuinely effective at forecasting startup success.</p><p>In another backtest, we trained the o1 reasoning model to assess and score historical YC startups across various dimensions of interest to early-stage investors like us. This is very similar to how human investors think through whether to invest in a given company.</p><p>As you&#8217;ll see in the chart below, not only were the reasoning model&#8217;s overall scores more predictive of startup success than our internal partner ratings, but some individual scoring dimensions like team quality, rate of progress, and product strength were <em>much</em> more predictive. As an experienced VC with 300+ startup investments under my belt, I noticed the dimensions where the model ratings were most predictive align perfectly with those that I&#8217;ve learned over the years to weigh more heavily when making investment decisions.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*z_jxeP6Ka5wm7t6gyOXQqw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*z_jxeP6Ka5wm7t6gyOXQqw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*z_jxeP6Ka5wm7t6gyOXQqw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*z_jxeP6Ka5wm7t6gyOXQqw.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*z_jxeP6Ka5wm7t6gyOXQqw.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*z_jxeP6Ka5wm7t6gyOXQqw.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*z_jxeP6Ka5wm7t6gyOXQqw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*z_jxeP6Ka5wm7t6gyOXQqw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*z_jxeP6Ka5wm7t6gyOXQqw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*z_jxeP6Ka5wm7t6gyOXQqw.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Now that I&#8217;ve hopefully got your attention on just how powerful these new AI reasoning models are, let&#8217;s talk about the implications for early-stage venture capital investors in the near future.</p><ol><li><p><strong>VCs can no longer afford </strong><em><strong>not</strong></em><strong> to use AI in their screening decisions</strong></p></li></ol><p>I&#8217;ve long felt that using data, technology, and algorithms has given us an edge at Rebel, but I now believe that using them are table stakes for VCs. We already live in a world where the latest AI models, if used properly, can make better startup success predictions than not only the average VC, but the <em>very best</em> VCs, and that gap will only widen as the underlying models evolve, and we get better at model data preparation, fine-tuning, prompting, etc. Remember, this is the worst these models will ever be.</p><p>This is not to suggest that VCs should rely entirely on AI models in making their investment decisions&#8202;&#8212;&#8202;at least not yet. At Rebel, we use ML/AI heavily in our startup screening process, but every investment decision is still ultimately made by a partner, as there are still nuances that AI can&#8217;t assess. We&#8217;ve been slowly incorporating more data signals into our algorithms as we backtest and validate them, and slowly handing over more influence to our ML/AI models as they prove themselves better at certain aspects of diligence than humans, and will continue doing so indefinitely.</p><p><strong>2. VCs must start viewing ourselves as technologists, not just investors</strong></p><p>Most venture funds view internal use of technology as an afterthought, despite our universal conviction that technology is the future. There&#8217;s an irony in manually drafting an investment memo for a cutting-edge AI startup investment in Microsoft Word (at Rebel, we use AI for that too)</p><p>Venture investing is often viewed today as an artisanal craft, requiring years of apprenticeship to master. That was true before the advent of advanced AI, but it&#8217;s not true today. VCs should stop thinking of ourselves as trained masters of our craft with impeccable tastes, but rather as orchestrators of a technology that&#8217;s far smarter than we are.</p><p>At Rebel, we&#8217;re uniquely situated to do this for YC startup investing with the most comprehensive dataset that exists of YC companies, founders, and outcomes for model training, plus a top-notch data science and software engineering team to make use of it&#8202;&#8212;&#8202;but <em>every</em> fund should hire at least one engineer and start training, backtesting, and using their own models based on their own data and investment strategy.</p><p><strong>3. LPs need to start asking fund managers different questions</strong></p><p>I&#8217;ve raised hundreds of millions of dollars from LPs over the years, so have a pretty good feel for the types of questions they ask venture fund managers. These questions tend to cover some combination of fund thesis, partnership team, portfolio strategy, track record, deal selection process, access advantages, GP experience/network, and portfolio support.</p><p>These all are relevant topics to cover, but noticeably absent are questions around data &amp; technology. In a world where advanced AI can outperform the best human investors in many regards, LPs should start getting comfortable asking questions like:</p><ul><li><p>How do you use ML/AI in your screening and due diligence process?</p></li><li><p>What is the background of your data &amp; software engineering team?</p></li><li><p>How much do you spend on AI inference and which models do you use?</p></li><li><p>How would you describe your data automation infrastructure?</p></li><li><p>How much have you invested into R&amp;D since fund inception?</p></li></ul><p>The top-performing venture funds in the world of AI won&#8217;t just invest in cutting-edge technology companies, they will <em>be</em> cutting-edge technology companies, and LPs will need to learn to diligence them as such. In this strange new world, a manager&#8217;s investing track record from 5+ years ago won&#8217;t be nearly as predictive of their future success as how well they&#8217;re harnessing AI technology today, at least in terms of startup due diligence and deal selection (though overall fund strategy, GP networks, deal access, and portfolio support certainly still matter)</p><p>I&#8217;ll conclude with one final thought for my fellow early-stage venture investors who remain skeptical: <em><strong>If you believe AI will revolutionize every industry, why wouldn&#8217;t it revolutionize ours?</strong></em></p>]]></content:encoded></item><item><title><![CDATA[On the last decade of Y Combinator]]></title><description><![CDATA[Since taking my last startup through Y Combinator as a founder over a decade ago, I&#8217;ve had the privilege of attending every YC Demo Day since my own in 2013 and investing in every YC batch professionally since 2018.]]></description><link>https://insights.rebelfund.vc/p/on-the-last-decade-of-y-combinator-fa19b387846a</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-the-last-decade-of-y-combinator-fa19b387846a</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Tue, 18 Feb 2025 23:10:27 GMT</pubDate><enclosure url="https://cdn-images-1.medium.com/max/800/1*7IrwcFPLbzok9mEGDWL9RQ.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>Since taking my last startup through Y Combinator as a founder over a decade ago, I&#8217;ve had the privilege of attending every YC Demo Day since my own in 2013 and investing in every YC batch professionally since 2018. It&#8217;s been fun to watch YC grow and evolve over the years, transitioning from Paul Graham&#8217;s leadership (2005&#8211;2014) to Sam Altman&#8217;s (2014&#8211;2019) to Geoff Ralston&#8217;s (2019&#8211;2022) to Garry Tan&#8217;s (2023-present). YC has really come into its own as an institution and had an enormous impact in Silicon Valley.</p><p>At <a href="https://www.rebelfund.vc/">Rebel Fund</a>, we&#8217;re positioned to be the single largest supporter of YC startups at the seed stage, now with close to $200M committed to invest in YC startups across our various funds, and portfolio companies nowvalued at over $12B in aggregate and growing. As an extremely data-driven fund entirely focused on the YC ecosystem, we&#8217;ve also built the world&#8217;s largest database of YC startups and founders outside of YC itself, now with millions of data points on thousands of companies, which puts us in a great position to track YC startup trends over time.</p><p>This post will highlight some of the trends we&#8217;ve discovered in YC companies and founders over the past decade through the lens of our data.</p><h3>Companies</h3><p>I&#8217;ll start with the most easily observable trend&#8202;&#8212;&#8202;number of companies by batch year. YC had been ramping up batch sizes steadily since 2015, but it really accelerated in the Covid / ZIRP years, in part because the program went 100% virtual. Lately things have settled at around 500&#8211;600 startups per year, though I expect this number to start growing again soon thanks to the AI revolution and YC moving from 2x to 4x batches per year.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*7IrwcFPLbzok9mEGDWL9RQ.png 424w, https://cdn-images-1.medium.com/max/800/1*7IrwcFPLbzok9mEGDWL9RQ.png 848w, https://cdn-images-1.medium.com/max/800/1*7IrwcFPLbzok9mEGDWL9RQ.png 1272w, https://cdn-images-1.medium.com/max/800/1*7IrwcFPLbzok9mEGDWL9RQ.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*7IrwcFPLbzok9mEGDWL9RQ.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*7IrwcFPLbzok9mEGDWL9RQ.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*7IrwcFPLbzok9mEGDWL9RQ.png 424w, https://cdn-images-1.medium.com/max/800/1*7IrwcFPLbzok9mEGDWL9RQ.png 848w, https://cdn-images-1.medium.com/max/800/1*7IrwcFPLbzok9mEGDWL9RQ.png 1272w, https://cdn-images-1.medium.com/max/800/1*7IrwcFPLbzok9mEGDWL9RQ.png 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p>Whenever I mention YC&#8217;s growing batch sizes to LPs, the immediate question they ask is some version of &#8220;Don&#8217;t you worry that the quality of YC startups is getting diluted?&#8221;</p><p>This concern is amplified by the fact that most of YC&#8217;s most famous and valuable startups were founded many years ago, giving the impression that YC&#8217;s best days are behind it. Our own data indeed shows a downward trend over the past decade in the proportion of YC startups that have been acquired or gone public, and those that have achieved unicorn ($1B+) or top companies ($150M+) valuations:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*8W1vNGP_uX3wsvb5O-Ea7Q.png 424w, https://cdn-images-1.medium.com/max/800/1*8W1vNGP_uX3wsvb5O-Ea7Q.png 848w, https://cdn-images-1.medium.com/max/800/1*8W1vNGP_uX3wsvb5O-Ea7Q.png 1272w, https://cdn-images-1.medium.com/max/800/1*8W1vNGP_uX3wsvb5O-Ea7Q.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*8W1vNGP_uX3wsvb5O-Ea7Q.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*8W1vNGP_uX3wsvb5O-Ea7Q.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*8W1vNGP_uX3wsvb5O-Ea7Q.png 424w, https://cdn-images-1.medium.com/max/800/1*8W1vNGP_uX3wsvb5O-Ea7Q.png 848w, https://cdn-images-1.medium.com/max/800/1*8W1vNGP_uX3wsvb5O-Ea7Q.png 1272w, https://cdn-images-1.medium.com/max/800/1*8W1vNGP_uX3wsvb5O-Ea7Q.png 1456w" sizes="100vw"></picture><div></div></div></div></figure></div><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*r5a9bC8jDRnNIzOQXMxXKQ.png 424w, https://cdn-images-1.medium.com/max/800/1*r5a9bC8jDRnNIzOQXMxXKQ.png 848w, https://cdn-images-1.medium.com/max/800/1*r5a9bC8jDRnNIzOQXMxXKQ.png 1272w, https://cdn-images-1.medium.com/max/800/1*r5a9bC8jDRnNIzOQXMxXKQ.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*r5a9bC8jDRnNIzOQXMxXKQ.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*r5a9bC8jDRnNIzOQXMxXKQ.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*r5a9bC8jDRnNIzOQXMxXKQ.png 424w, https://cdn-images-1.medium.com/max/800/1*r5a9bC8jDRnNIzOQXMxXKQ.png 848w, https://cdn-images-1.medium.com/max/800/1*r5a9bC8jDRnNIzOQXMxXKQ.png 1272w, https://cdn-images-1.medium.com/max/800/1*r5a9bC8jDRnNIzOQXMxXKQ.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Of course, what these charts fail to take into account is it takes <a href="https://link.medium.com/QR3zTM87Oub">7 years on average</a> for a YC startup to achieve an exit, so naturally the most successful startups were founded a long time ago.</p><p>The good news is every leading indicator of YC batch quality that we track at Rebel continues improving over time despite growing batch sizes, from YC application acceptance rates to fundraising progress to founder quality metrics. So, we believe YC&#8217;s future success will far outpace the outcomes it&#8217;s achieved thus far.</p><p>The composition of YC batches has also evolved over time. While B2B startups have always been a mainstay, consumer companies have definitely fallen out of favor in recent years, and fintech experienced a resurgence in the ZIRP years (logically). In the last couple of years that have kicked off the AI revolution, B2B is stronger than ever, as AI-native vertical SaaS and dev tool startups are creating huge opportunity for founders and investors.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*kYK2_-qgDN0AcPFcGXfKrA.png 424w, https://cdn-images-1.medium.com/max/800/1*kYK2_-qgDN0AcPFcGXfKrA.png 848w, https://cdn-images-1.medium.com/max/800/1*kYK2_-qgDN0AcPFcGXfKrA.png 1272w, https://cdn-images-1.medium.com/max/800/1*kYK2_-qgDN0AcPFcGXfKrA.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*kYK2_-qgDN0AcPFcGXfKrA.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*kYK2_-qgDN0AcPFcGXfKrA.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*kYK2_-qgDN0AcPFcGXfKrA.png 424w, https://cdn-images-1.medium.com/max/800/1*kYK2_-qgDN0AcPFcGXfKrA.png 848w, https://cdn-images-1.medium.com/max/800/1*kYK2_-qgDN0AcPFcGXfKrA.png 1272w, https://cdn-images-1.medium.com/max/800/1*kYK2_-qgDN0AcPFcGXfKrA.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>All technology sectors are not created equally though&#8202;&#8212;&#8202;our data shows that it&#8217;s much easier to build a top company or unicorn in some areas than others. Fintech is a standout on the &#8216;easier&#8217; side, and apparently it&#8217;s impossible to build a unicorn serving the Government &#129300;</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*3XLVzyCyeulyKxygC39T8w.png 424w, https://cdn-images-1.medium.com/max/800/1*3XLVzyCyeulyKxygC39T8w.png 848w, https://cdn-images-1.medium.com/max/800/1*3XLVzyCyeulyKxygC39T8w.png 1272w, https://cdn-images-1.medium.com/max/800/1*3XLVzyCyeulyKxygC39T8w.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*3XLVzyCyeulyKxygC39T8w.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*3XLVzyCyeulyKxygC39T8w.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*3XLVzyCyeulyKxygC39T8w.png 424w, https://cdn-images-1.medium.com/max/800/1*3XLVzyCyeulyKxygC39T8w.png 848w, https://cdn-images-1.medium.com/max/800/1*3XLVzyCyeulyKxygC39T8w.png 1272w, https://cdn-images-1.medium.com/max/800/1*3XLVzyCyeulyKxygC39T8w.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>YC startups have also waxed and waned by geography. In the early years, YC was almost entirely US-focused, but it became much more receptive to international startups (particularly from emerging markets) during the Covid / ZIRP years, but has recently refocused squarely on the US and other developed markets. My guess is YC partners saw the same trend in their data that Rebel saw in ours&#8230; the closer a startup gets to Silicon Valley, the more likely they are to become a unicorn.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*HemX7Tjfh3CplfW7kqLFQg.png 424w, https://cdn-images-1.medium.com/max/800/1*HemX7Tjfh3CplfW7kqLFQg.png 848w, https://cdn-images-1.medium.com/max/800/1*HemX7Tjfh3CplfW7kqLFQg.png 1272w, https://cdn-images-1.medium.com/max/800/1*HemX7Tjfh3CplfW7kqLFQg.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*HemX7Tjfh3CplfW7kqLFQg.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*HemX7Tjfh3CplfW7kqLFQg.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*HemX7Tjfh3CplfW7kqLFQg.png 424w, https://cdn-images-1.medium.com/max/800/1*HemX7Tjfh3CplfW7kqLFQg.png 848w, https://cdn-images-1.medium.com/max/800/1*HemX7Tjfh3CplfW7kqLFQg.png 1272w, https://cdn-images-1.medium.com/max/800/1*HemX7Tjfh3CplfW7kqLFQg.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>An interesting quirk we noticed is that YC&#8217;s Winter batches (January&#8202;&#8212;&#8202;March) seem to outperform its Summer batches (June&#8202;&#8212;&#8202;August). The Winter batches are about ~10% larger on average, but notably, they raised nearly 70% of all the capital! Maybe in the summertime, VC&#8217;s are too busy galavanting across Europe, kitesurfing in Puerto Rico, or partying at Burning Man to invest in startups (those who know me will appreciate how self-deprecating that was &#128514;)</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*mq9S77wAeB3J6ivQY1Avhw.png 424w, https://cdn-images-1.medium.com/max/800/1*mq9S77wAeB3J6ivQY1Avhw.png 848w, https://cdn-images-1.medium.com/max/800/1*mq9S77wAeB3J6ivQY1Avhw.png 1272w, https://cdn-images-1.medium.com/max/800/1*mq9S77wAeB3J6ivQY1Avhw.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*mq9S77wAeB3J6ivQY1Avhw.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*mq9S77wAeB3J6ivQY1Avhw.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*mq9S77wAeB3J6ivQY1Avhw.png 424w, https://cdn-images-1.medium.com/max/800/1*mq9S77wAeB3J6ivQY1Avhw.png 848w, https://cdn-images-1.medium.com/max/800/1*mq9S77wAeB3J6ivQY1Avhw.png 1272w, https://cdn-images-1.medium.com/max/800/1*mq9S77wAeB3J6ivQY1Avhw.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><h3>Founders</h3><p>YC startups are nothing without their founders, and since founder quality is the #1 factor we underwrite at Rebel, the vast majority of data we collect and feed into our <a href="https://link.medium.com/lPJ0A1Zc4Kb">Rebel Theorem 3.0 machine learning algorithm</a> is founder-related. There are lots of trends I could speak to regarding YC founders, but here I&#8217;ll highlight a few of the more interesting ones.</p><p>Let&#8217;s first talk about YC founders&#8217; age, which we don&#8217;t track explicitly at Rebel, but can infer by how many years passed between when they started college and their YC batch (we used to look at years since they finished college, but realized that doesn&#8217;t account for dropouts). The chart below shows &#8220;school to batch&#8221; historically averaged around 9 years, which implies an average YC founder age of 27, assuming they started college at 18.</p><p>What&#8217;s interesting is average YC founder age has reached historic lows in the past couple of years, hitting around 25 years in 2024. I believe this is a direct result of the AI revolution&#8202;&#8212;&#8202;since generative AI is such a new technology, younger founders are at no disadvantage understanding and building around it, so YC batches are trending younger.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*LJ0Vx9i0H44e5HN40q1VOw.png 424w, https://cdn-images-1.medium.com/max/800/1*LJ0Vx9i0H44e5HN40q1VOw.png 848w, https://cdn-images-1.medium.com/max/800/1*LJ0Vx9i0H44e5HN40q1VOw.png 1272w, https://cdn-images-1.medium.com/max/800/1*LJ0Vx9i0H44e5HN40q1VOw.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*LJ0Vx9i0H44e5HN40q1VOw.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*LJ0Vx9i0H44e5HN40q1VOw.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*LJ0Vx9i0H44e5HN40q1VOw.png 424w, https://cdn-images-1.medium.com/max/800/1*LJ0Vx9i0H44e5HN40q1VOw.png 848w, https://cdn-images-1.medium.com/max/800/1*LJ0Vx9i0H44e5HN40q1VOw.png 1272w, https://cdn-images-1.medium.com/max/800/1*LJ0Vx9i0H44e5HN40q1VOw.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Another way we look at YC founder experience at Rebel is how many years of experience founders had <em>as a startup co-founder</em>. Here we see an even sharper drop-off in the last couple of years, with average co-founder experience hovering around 3 years for the most of the past decade, but declining to half of that last year. The most recent cohorts of mostly AI-focused YC founders are much less experienced as startup founders.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*9CaOPhut72TgKFzix8Tj5g.png 424w, https://cdn-images-1.medium.com/max/800/1*9CaOPhut72TgKFzix8Tj5g.png 848w, https://cdn-images-1.medium.com/max/800/1*9CaOPhut72TgKFzix8Tj5g.png 1272w, https://cdn-images-1.medium.com/max/800/1*9CaOPhut72TgKFzix8Tj5g.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*9CaOPhut72TgKFzix8Tj5g.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*9CaOPhut72TgKFzix8Tj5g.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*9CaOPhut72TgKFzix8Tj5g.png 424w, https://cdn-images-1.medium.com/max/800/1*9CaOPhut72TgKFzix8Tj5g.png 848w, https://cdn-images-1.medium.com/max/800/1*9CaOPhut72TgKFzix8Tj5g.png 1272w, https://cdn-images-1.medium.com/max/800/1*9CaOPhut72TgKFzix8Tj5g.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>They&#8217;re also much more likely to be first-time founders&#8230;</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*VrqHyd4y7CTK_vaLy-g7FQ.png 424w, https://cdn-images-1.medium.com/max/800/1*VrqHyd4y7CTK_vaLy-g7FQ.png 848w, https://cdn-images-1.medium.com/max/800/1*VrqHyd4y7CTK_vaLy-g7FQ.png 1272w, https://cdn-images-1.medium.com/max/800/1*VrqHyd4y7CTK_vaLy-g7FQ.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*VrqHyd4y7CTK_vaLy-g7FQ.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*VrqHyd4y7CTK_vaLy-g7FQ.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*VrqHyd4y7CTK_vaLy-g7FQ.png 424w, https://cdn-images-1.medium.com/max/800/1*VrqHyd4y7CTK_vaLy-g7FQ.png 848w, https://cdn-images-1.medium.com/max/800/1*VrqHyd4y7CTK_vaLy-g7FQ.png 1272w, https://cdn-images-1.medium.com/max/800/1*VrqHyd4y7CTK_vaLy-g7FQ.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>&#8230;and much more likely to be technical:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*rP3WGKQHpU6YOtEhK7bfOw.png 424w, https://cdn-images-1.medium.com/max/800/1*rP3WGKQHpU6YOtEhK7bfOw.png 848w, https://cdn-images-1.medium.com/max/800/1*rP3WGKQHpU6YOtEhK7bfOw.png 1272w, https://cdn-images-1.medium.com/max/800/1*rP3WGKQHpU6YOtEhK7bfOw.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*rP3WGKQHpU6YOtEhK7bfOw.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*rP3WGKQHpU6YOtEhK7bfOw.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*rP3WGKQHpU6YOtEhK7bfOw.png 424w, https://cdn-images-1.medium.com/max/800/1*rP3WGKQHpU6YOtEhK7bfOw.png 848w, https://cdn-images-1.medium.com/max/800/1*rP3WGKQHpU6YOtEhK7bfOw.png 1272w, https://cdn-images-1.medium.com/max/800/1*rP3WGKQHpU6YOtEhK7bfOw.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>The YC founder characteristics we&#8217;ve covered so far are relatively straightforward and easy to measure. However, we also look at some less obvious founder characteristics at Rebel, like founder personalities types as assessed by their online footprint using AI.</p><p>The biggest shift we&#8217;ve noticed is that YC founder Dominance and Influence personality traits have decreased dramatically over the past couple of years, which makes sense given how much younger and more technical the founders are these days.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/0*EQo3SndGhHHbDwvx 424w, https://cdn-images-1.medium.com/max/800/0*EQo3SndGhHHbDwvx 848w, https://cdn-images-1.medium.com/max/800/0*EQo3SndGhHHbDwvx 1272w, https://cdn-images-1.medium.com/max/800/0*EQo3SndGhHHbDwvx 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/0*EQo3SndGhHHbDwvx" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/0*EQo3SndGhHHbDwvx&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/0*EQo3SndGhHHbDwvx 424w, https://cdn-images-1.medium.com/max/800/0*EQo3SndGhHHbDwvx 848w, https://cdn-images-1.medium.com/max/800/0*EQo3SndGhHHbDwvx 1272w, https://cdn-images-1.medium.com/max/800/0*EQo3SndGhHHbDwvx 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>These younger and more technical founders are more likely to display Steadiness and Conscientiousness personality traits, which also goes with the &#8220;nerdy engineer&#8221; stereotype.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/0*JloG92e5XQhX2BBX 424w, https://cdn-images-1.medium.com/max/800/0*JloG92e5XQhX2BBX 848w, https://cdn-images-1.medium.com/max/800/0*JloG92e5XQhX2BBX 1272w, https://cdn-images-1.medium.com/max/800/0*JloG92e5XQhX2BBX 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/0*JloG92e5XQhX2BBX" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/0*JloG92e5XQhX2BBX&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/0*JloG92e5XQhX2BBX 424w, https://cdn-images-1.medium.com/max/800/0*JloG92e5XQhX2BBX 848w, https://cdn-images-1.medium.com/max/800/0*JloG92e5XQhX2BBX 1272w, https://cdn-images-1.medium.com/max/800/0*JloG92e5XQhX2BBX 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>There&#8217;s a lot more I could say about YC founder personality traits, how we measure them at Rebel, and how predictive they are of startup success, but I&#8217;ll save that for another blog post.</p><p>YC startup and founder profiles have and certainly will continue to evolve with the changing technological and macroeconomic landscape. You can probably infer a lot about the broader technology world by looking at these YC trends as well, so I hope you&#8217;ve found at least some of these insightful!</p>]]></content:encoded></item><item><title><![CDATA[On seed-stage startup traction (and why to ignore it)]]></title><description><![CDATA[I&#8217;ve had the privilege of investing in 300+ Y Combinator startups over my career and looked at many thousands more.]]></description><link>https://insights.rebelfund.vc/p/on-seed-stage-startup-traction-and-why-to-ignore-it-9ad662981145</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-seed-stage-startup-traction-and-why-to-ignore-it-9ad662981145</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Fri, 07 Feb 2025 17:08:13 GMT</pubDate><enclosure url="https://cdn-images-1.medium.com/max/600/1*EZB5Ww1WCAMMz2alESyUIQ.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>I&#8217;ve had the privilege of investing in 300+ Y Combinator startups over my career and looked at many thousands more. Over the years, I&#8217;ve learned that one of the trickiest things about seed-stage investing is separating &#8220;signal&#8221; from &#8220;noise&#8221;. Every company, team, idea, product, and market is unique and exciting, but a good VC has to know what information to pay attention to and what to ignore when making their investment decisions.</p><p>At <a href="https://www.rebelfund.vc/">Rebel Fund</a>, we&#8217;ve taken a very data-driven and scientific approach to separating signal and noise when looking at YC startups. We&#8217;ve invested millions of dollars into developing the world&#8217;s most comprehensive database of YC startups and founders outside of YC itself, now tracking hundreds of variables on each company and millions of data points overall. We&#8217;ve used this data to train our proprietary <a href="https://link.medium.com/lPJ0A1Zc4Kb">Rebel Theorem 3.0 machine learning algorithm</a>, which scores every new YC startup in terms of its likelihood of having a successful outcome.</p><p>When making our investment decisions, most VCs look at some combination of the same aspects of each company: team, product, market, traction, timing, competition, and defensibility. At Rebel, we pay attention to all of these things at different stages of our due diligence process, but focus our algorithm most heavily on characteristics (or in data science speak &#8220;features&#8221;) related to the founders themselves. This is because product ideas can and often do pivot over time, markets rapidly evolve, traction can wane, competitors come and go&#8230; but the founders are usually a constant and have a huge impact on the fate of an early-stage startup.</p><p>However, many seed-stage investors prefer to focus on the more easily quantifiable aspects of a startup when making their investment decisions, like revenue traction or monthly growth rate, probably because it feels more comfortable to base a high-stakes investment decision on something &#8216;hard&#8217; like numbers vs something &#8216;soft&#8217; like team quality. This may be why almost every YC startup has an &#8220;up and to the right&#8221; chart in their pitch deck, showing how their revenue is growing like crazy (though often with just a few weeks or months worth of history).</p><p>This begs a very important question, which I hope to anwer in this post: <em><strong>From an investor standpoint, how much does early traction really matter in seed-stage startups?</strong></em></p><p>I asked this question to several YC partners I know, since arguably no one in the world is in a better position to pattern-match how much early startup traction predicts their long-term success. Shockingly, I found no consensus whatsoever. Some partners felt it was super important, others felt it doesn&#8217;t matter at all, and others felt it matters but with certain qualifiers, like only above a certain revenue threshold, or only in the sense that it shows founders have hustle or sales skills.</p><p>I found it curious there&#8217;s an aspect of seed-stage startups that investors seem to care about very deeply, yet even YC partners can&#8217;t seem to agree whether it actually matters.</p><p>So, I did what any tech nerd armed with a massive database and top-notch data science team would do&#8202;&#8212;&#8202;set out to answer this question quantitatively.</p><p>Rebel has been collecting early traction metrics on YC startups at Demo Day since early 2022 (I wish we had started earlier&#185;). Here are the ones we determined most relevant for this analysis:</p><p><strong>Demo MRR</strong>&#8202;&#8212;&#8202;The monthly recurring revenue (MRR) reported by startups at Demo Day. When revenue was reported weekly or annualized, we standardize it to monthly.</p><p><strong>Demo Monthly Growth Rate&#8202;</strong>&#8212;&#8202;The month-over-month (MoM) growth rate reported by startups at Demo Day. When reported weekly or annualized, we also standardize it to monthly.</p><p><strong>Demo Total LOI</strong>&#8202;&#8212;&#8202;The total dollar amount of revenue pre-committed to startups via a letter of intent (LOI) as reported at Demo Day. These are non-binding commitment letters, typically used by capital-intensive startups to show market interest in their product ahead of commercializing it.</p><p><strong>Demo Annualized GMV</strong>&#8202;&#8212;&#8202;The annualized gross merchandise volume (GMV) reported by startups at Demo Day. This metric is typically used by marketplace startups to show the value of products or services passing through their platform.</p><p>Of course, not all of these metrics are reported or even relevant to all startups, but we still had thousands of data points to work with.</p><p>We then compared these early traction metrics to an internal metric we use called <strong>Latest Inferred Valuation Final</strong>, which is our best estimate of the latest valuation of each startup, based on a proprietary formula we developed incorporating various internal and external data sources.</p><p>Our data analyst then ran a correlation analysis to see how well these early traction metrics correlate with companies&#8217; latest valuation. If indeed strong early traction is highly correlated (and plausibly predictive) of startup success, then we&#8217;d expect to see high correlation coefficients between at least some of these traction metrics and latest inferred valuations.</p><p>Below are correlation heatmaps our data analyst produced, using both the Spearman and Pearson methods (explaining these goes beyond the scope of this post). Here are some correlation strength guidelines for context:</p><ul><li><p>0.00 to &#177;0.10: Negligible correlation (no meaningful relationship)</p></li><li><p>&#177;0.10 to &#177;0.30: Weak correlation (small, but noticeable relationship)</p></li><li><p>&#177;0.30 to &#177;0.50: Moderate correlation (clear relationship, but not strong)</p></li><li><p>&#177;0.50 to &#177;1.00: Strong correlation (variables are strongly associated)</p></li><li><p>&#177;0.70 to &#177;1.00: Very strong correlation</p></li></ul><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/600/1*EZB5Ww1WCAMMz2alESyUIQ.png 424w, https://cdn-images-1.medium.com/max/600/1*EZB5Ww1WCAMMz2alESyUIQ.png 848w, https://cdn-images-1.medium.com/max/600/1*EZB5Ww1WCAMMz2alESyUIQ.png 1272w, https://cdn-images-1.medium.com/max/600/1*EZB5Ww1WCAMMz2alESyUIQ.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/600/1*EZB5Ww1WCAMMz2alESyUIQ.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/600/1*EZB5Ww1WCAMMz2alESyUIQ.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/600/1*EZB5Ww1WCAMMz2alESyUIQ.png 424w, https://cdn-images-1.medium.com/max/600/1*EZB5Ww1WCAMMz2alESyUIQ.png 848w, https://cdn-images-1.medium.com/max/600/1*EZB5Ww1WCAMMz2alESyUIQ.png 1272w, https://cdn-images-1.medium.com/max/600/1*EZB5Ww1WCAMMz2alESyUIQ.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/600/1*QcUI94yetNMcOaxvdzjYfg.png 424w, https://cdn-images-1.medium.com/max/600/1*QcUI94yetNMcOaxvdzjYfg.png 848w, https://cdn-images-1.medium.com/max/600/1*QcUI94yetNMcOaxvdzjYfg.png 1272w, https://cdn-images-1.medium.com/max/600/1*QcUI94yetNMcOaxvdzjYfg.png 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/600/1*QcUI94yetNMcOaxvdzjYfg.png" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/600/1*QcUI94yetNMcOaxvdzjYfg.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/600/1*QcUI94yetNMcOaxvdzjYfg.png 424w, https://cdn-images-1.medium.com/max/600/1*QcUI94yetNMcOaxvdzjYfg.png 848w, https://cdn-images-1.medium.com/max/600/1*QcUI94yetNMcOaxvdzjYfg.png 1272w, https://cdn-images-1.medium.com/max/600/1*QcUI94yetNMcOaxvdzjYfg.png 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>As you&#8217;ll see, <em><strong>there is a weak to negligible correlation between any of these traction metrics and latest company valuations.</strong></em></p><p>The only metric that makes a decent showing is <strong>Demo Total LOI</strong>, but I suspect that&#8217;s because LOIs are collected by capital-intensive startups (like hardware companies) that tend to raise at higher valuations by nature.</p><p>The implication is all of those &#8220;up and to the right&#8221; growth charts in YC startup pitch decks are mostly just noise from an investor perspective. This would also explain why the YC partners I spoke with had such wildly different opinions on whether early YC startup traction really matters.</p><p>This seems counter-intuitive, since isn&#8217;t the #1 job of every startup <em>to grow</em>?</p><p>It is indeed, but there&#8217;s a big difference between post-PMF (product-market fit) growth and initial pre-PMF growth just a few weeks or months after going live. The former is a vital prerequisite of a successful startup, and growth-stage investors should pay attention to it. However the latter is signal-poor and frankly easily manipulable by founders&#8202;&#8212;&#8202;it&#8217;s not hard to engineer a few weeks or months of growth by pumping up paid SEO or turning all your attention to sales ahead of Demo Day&#178;.</p><p>Even with this blog post in the wild, I&#8217;m sure that YC founders will continue to ramp up their sales ahead of Demo Day and investors will continue to swoon over their steep early growth curves. Savvy investors beware though&#8230; early traction in seed-stage startups is mostly just window dressing, so best to focus on the &#8216;soft&#8217; stuff &#128077;&#127995;</p><p>***************************************</p><p>&#185;I know ~3 years is a relatively short amount of time to judge the success of a startup, but it&#8217;s all we can do based on our data. It&#8217;s probably enough for this analysis anyway, since if early and sustained growth was the reason for a startup&#8217;s success, we&#8217;d likely see valuation markups pretty quickly.</p><p>&#178;From a <em>founder</em> perspective I think early traction still matters, because it forces them to speak with customers, deeply understand their target market, and evolve their product enough to earn real revenue dollars. And if it helps you raise capital, go for it!</p>]]></content:encoded></item><item><title><![CDATA[On YC unicorn founders]]></title><description><![CDATA[I&#8217;ve published dozens of data-driven blog posts over the years about Y Combinator startups and founders from an investor perspective, but I&#8217;m yet to talk about the people that made YC&#8217;s incredible success possible in the first place &#8212; its unicorn founders.]]></description><link>https://insights.rebelfund.vc/p/on-yc-unicorn-founders-6530aad4bb89</link><guid isPermaLink="false">https://insights.rebelfund.vc/p/on-yc-unicorn-founders-6530aad4bb89</guid><dc:creator><![CDATA[Jared Heyman]]></dc:creator><pubDate>Thu, 22 Aug 2024 15:23:13 GMT</pubDate><enclosure url="https://cdn-images-1.medium.com/max/800/1*R01NwKA-YJh9K3j1UwnwGw.jpeg" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>I&#8217;ve published <a href="https://jaredheyman.medium.com/">dozens of data-driven blog posts</a> over the years about <a href="https://www.ycombinator.com/">Y Combinator</a> startups and founders from an investor perspective, but I&#8217;m yet to talk about the people that made YC&#8217;s incredible success possible in the first place&#8202;&#8212;&#8202;its unicorn founders.</p><p>As I explained in <a href="https://link.medium.com/D6Ootw8LdJb">On the power law of Y Combinator startups</a>, the unicorns ($1B+ companies) drive the overwhelming majority of YC&#8217;s early-stage investor returns. Since at <a href="https://www.rebelfund.vc/">Rebel Fund</a>, we believe that a company&#8217;s founders are the ultimate drivers of its value creation, it only seems fair to take a deep dive into what YC unicorn founders &#8220;look like.&#8221;</p><p>Well, according to DALL-E based on some of our summary statistics, a typical YC unicorn founder looks like this:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*R01NwKA-YJh9K3j1UwnwGw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*R01NwKA-YJh9K3j1UwnwGw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*R01NwKA-YJh9K3j1UwnwGw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*R01NwKA-YJh9K3j1UwnwGw.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*R01NwKA-YJh9K3j1UwnwGw.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*R01NwKA-YJh9K3j1UwnwGw.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*R01NwKA-YJh9K3j1UwnwGw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*R01NwKA-YJh9K3j1UwnwGw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*R01NwKA-YJh9K3j1UwnwGw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*R01NwKA-YJh9K3j1UwnwGw.jpeg 1456w" sizes="100vw" fetchpriority="high"></picture><div></div></div></div></figure></div><p>Rather than stop the post here though, I&#8217;ll share a few highlights from the thousands of data points Rebel has collected and analyzed on the founders of the ~100 unicorns that YC has helped launch thus far.</p><p><strong>Experience</strong></p><p>YC unicorn founders had an average of <strong>8 years work experience</strong> when they started their YC startup. Assuming they got their first job after graduating college at around 22 years of age, that means they were around 30 years old on average.</p><p>However, as you&#8217;ll see in the chart below, there is a wide range of years experience when they founded their YC startups. Plenty of them had more or less than this average, and a surprising number were only a few years into their career&#8202;&#8212;&#8202;though relatively few had over 15 years experience.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*Rb83UEnR7sdtp8eGKTuhsw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Rb83UEnR7sdtp8eGKTuhsw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Rb83UEnR7sdtp8eGKTuhsw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Rb83UEnR7sdtp8eGKTuhsw.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*Rb83UEnR7sdtp8eGKTuhsw.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*Rb83UEnR7sdtp8eGKTuhsw.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*Rb83UEnR7sdtp8eGKTuhsw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*Rb83UEnR7sdtp8eGKTuhsw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*Rb83UEnR7sdtp8eGKTuhsw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*Rb83UEnR7sdtp8eGKTuhsw.jpeg 1456w" sizes="100vw"></picture><div></div></div></div></figure></div><p>I&#8217;m often asked how many top YC founders are first-time vs repeat founders, and of course this is also something we track at Rebel. The chart below shows how many times each unicorn founder was a startup co-founder <em>before</em> their YC startup:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*BP1xNsVkvKH-g5N7ISCZeg.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*BP1xNsVkvKH-g5N7ISCZeg.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*BP1xNsVkvKH-g5N7ISCZeg.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*BP1xNsVkvKH-g5N7ISCZeg.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*BP1xNsVkvKH-g5N7ISCZeg.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*BP1xNsVkvKH-g5N7ISCZeg.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*BP1xNsVkvKH-g5N7ISCZeg.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*BP1xNsVkvKH-g5N7ISCZeg.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*BP1xNsVkvKH-g5N7ISCZeg.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*BP1xNsVkvKH-g5N7ISCZeg.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>As you&#8217;ll see, a slight majority of them were first-time founders, though a signficant number had co-founded one or more startups in the past. I believe this tells us that co-founder experience is helpful but not necessary to build a successful venture-backed startup.</p><p>Only a small minority of venture-backed startups achieve successful exits (acquistion or IPO) so perhaps it&#8217;s unsurprising that a smaller proportion of YC unicorn founders had lead a startup all the way to an acquisition prior to starting their YC company&#8202;&#8212;&#8202;about 20% of them.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*77V-QcE4WEgAHv2GnUtGBw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*77V-QcE4WEgAHv2GnUtGBw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*77V-QcE4WEgAHv2GnUtGBw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*77V-QcE4WEgAHv2GnUtGBw.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*77V-QcE4WEgAHv2GnUtGBw.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*77V-QcE4WEgAHv2GnUtGBw.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*77V-QcE4WEgAHv2GnUtGBw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*77V-QcE4WEgAHv2GnUtGBw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*77V-QcE4WEgAHv2GnUtGBw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*77V-QcE4WEgAHv2GnUtGBw.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p><strong>Education</strong></p><p>So now you may be wondering, where did these unicorn founders go to school? Well, most likely Stanford or MIT&nbsp;:-)</p><p>The chart below shows the number of YC unicorn founders with degrees at universities ranked by Rebel&#8217;s internal top school algorithm. Since YC has always had a bias towards technical founders, top technical universities are represented most often:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*mD1zsWWC3SLDmCoL1LMLaw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*mD1zsWWC3SLDmCoL1LMLaw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*mD1zsWWC3SLDmCoL1LMLaw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*mD1zsWWC3SLDmCoL1LMLaw.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*mD1zsWWC3SLDmCoL1LMLaw.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*mD1zsWWC3SLDmCoL1LMLaw.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*mD1zsWWC3SLDmCoL1LMLaw.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*mD1zsWWC3SLDmCoL1LMLaw.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*mD1zsWWC3SLDmCoL1LMLaw.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*mD1zsWWC3SLDmCoL1LMLaw.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>While a couple of elite engineering universities are seen most often on unicorn founders&#8217; resumes, there is a long tail of other universities represented, so going to a top engineering school also falls into the &#8220;helpful but not necessary&#8221; category for unicorn founders.</p><p>What did they study? As you may have guessed, probably something technical.</p><p>The chart below shows the degree types most often seen on YC unicorn founders&#8217; diplomas. Of course Bachelor of Science (BS) is most common, but there are also a lot of BAs and few MBAs, and plenty of post-graduate degrees in technical fields.</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*WfJa1IbgjRwIurWPJBAZ_g.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*WfJa1IbgjRwIurWPJBAZ_g.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*WfJa1IbgjRwIurWPJBAZ_g.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*WfJa1IbgjRwIurWPJBAZ_g.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*WfJa1IbgjRwIurWPJBAZ_g.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*WfJa1IbgjRwIurWPJBAZ_g.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*WfJa1IbgjRwIurWPJBAZ_g.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*WfJa1IbgjRwIurWPJBAZ_g.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*WfJa1IbgjRwIurWPJBAZ_g.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*WfJa1IbgjRwIurWPJBAZ_g.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p><strong>Demographics</strong></p><p>As I discussed in my post <a href="https://link.medium.com/lPJ0A1Zc4Kb">On Rebel Theorem 3.0</a>, our proprietary ML/AI algorithm for predicting YC startup success, we&#8217;ve observed that founder demographics (age, gender, ethnicity, etc) is not a reliable predictor of YC startup outcomes. That said, it&#8217;s worth noting a few demographic characteristics of YC unicorn founders.</p><p>The first is that they&#8217;re overwhelmingly male, at least as inferred by GPT-4 based on their first and last name:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*ZKImowdU8M2meirwjiz1jg.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*ZKImowdU8M2meirwjiz1jg.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*ZKImowdU8M2meirwjiz1jg.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*ZKImowdU8M2meirwjiz1jg.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*ZKImowdU8M2meirwjiz1jg.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*ZKImowdU8M2meirwjiz1jg.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*ZKImowdU8M2meirwjiz1jg.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*ZKImowdU8M2meirwjiz1jg.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*ZKImowdU8M2meirwjiz1jg.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*ZKImowdU8M2meirwjiz1jg.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>The second is that they&#8217;re quite commonly immigrants to the US. We also use GPT-4 to infer a founder&#8217;s most likely country of origen based on their educational and work experience, and we&#8217;ve found that while nearly 60% of YC unicorn founders are originally American, a significant minority are from outside the US:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*0wGyLpqeW0899qgwjOPxAA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*0wGyLpqeW0899qgwjOPxAA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*0wGyLpqeW0899qgwjOPxAA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*0wGyLpqeW0899qgwjOPxAA.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*0wGyLpqeW0899qgwjOPxAA.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*0wGyLpqeW0899qgwjOPxAA.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*0wGyLpqeW0899qgwjOPxAA.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*0wGyLpqeW0899qgwjOPxAA.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*0wGyLpqeW0899qgwjOPxAA.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*0wGyLpqeW0899qgwjOPxAA.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>As I discussed in my post <a href="https://link.medium.com/hHjsW0aAtub">On 101 Y Combinator unicorns</a>, almost all YC unicorn <em>companies</em> are in the US (with about 2/3 of those in the SF Bay Area alone) but that does <em>not</em> mean that the founders originated there.</p><p><strong>Personality</strong></p><p>Also as discussed in my post <a href="https://link.medium.com/lPJ0A1Zc4Kb">On Rebel Theorem 3.0</a>, we can now use AI to accurately infer a founder&#8217;s personality profile based on certain aspects of their online footprint. Our findings on YC founder personality profiles probably warrant their own blog post at some point, but for now let&#8217;s just look at how the DISC profile of YC unicorn founders compare to the general population:</p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://cdn-images-1.medium.com/max/800/1*GOeI67UpZBrIOj2BFsa14Q.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*GOeI67UpZBrIOj2BFsa14Q.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*GOeI67UpZBrIOj2BFsa14Q.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*GOeI67UpZBrIOj2BFsa14Q.jpeg 1456w" sizes="100vw"><img src="https://cdn-images-1.medium.com/max/800/1*GOeI67UpZBrIOj2BFsa14Q.jpeg" data-attrs="{&quot;src&quot;:&quot;https://cdn-images-1.medium.com/max/800/1*GOeI67UpZBrIOj2BFsa14Q.jpeg&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:null,&quot;width&quot;:null,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:null,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:null,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://cdn-images-1.medium.com/max/800/1*GOeI67UpZBrIOj2BFsa14Q.jpeg 424w, https://cdn-images-1.medium.com/max/800/1*GOeI67UpZBrIOj2BFsa14Q.jpeg 848w, https://cdn-images-1.medium.com/max/800/1*GOeI67UpZBrIOj2BFsa14Q.jpeg 1272w, https://cdn-images-1.medium.com/max/800/1*GOeI67UpZBrIOj2BFsa14Q.jpeg 1456w" sizes="100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>The &#8220;average&#8221; YC unicorn founder has the following DISC profile:</p><p><strong>Dominance (D): 36%<br>Influence (I): 15%<br>Steadiness (S): 11%<br>Conscientiousness (C): 37%</strong></p><p>The high <strong>Dominance</strong> means they&#8217;re results-oriented, confident, and decisive. They prefer to take control of situations, focus on achieving goals, and tend to be direct in their communication. They value efficiency and can sometimes be perceived as assertive or even aggressive in pursuit of objectives.</p><p>The high <strong>Conscientiousness</strong> means they place importance on accuracy, quality, and detailed work. They have a strong sense of responsibility and a preference for working within established structures or guidelines. They&#8217;re be methodical in their approach and prioritize correctness in their work.</p><p>The low <strong>Influence</strong> means they&#8217;re reserved when it comes to social interactions or influencing others. They prefer working independently or in smaller groups, and may not actively seek to be the center of attention. They rely more on facts and logic than on charm or persuasion.</p><p>The low <strong>Steadiness</strong> means they thrive in environments that are fast-paced and dynamic rather than stable or predictable. They&#8217;re open to change and comfortable with a certain level of uncertainty, preferring variety over routine.</p><p>As someone who&#8217;s invested in hundreds of top YC startups, I find none of this surpising. While of course founder personalities vary and there&#8217;s no single personality archetype of a successful founder, most of the best founders I&#8217;ve worked with can be described by these traits.</p><p>There&#8217;s a lot more I could share about YC unicorn founders from our data, but I think this is a good place to wrap up this post. In case you&#8217;re wondering if YC unicorn founders have different characteristics than YC founders in general, the answer is &#8220;yes&#8221; but that&#8217;s best explained in my post <a href="https://link.medium.com/lPJ0A1Zc4Kb">On Rebel Theorem 3.0</a>.</p><p>If you&#8217;re an investor, happy unicorn hunting!</p>]]></content:encoded></item></channel></rss>