VCs Score Founders: Top-Fund Experience Ranks First, MBA at the Bottom, PhDs Just Turned the Corner

Deedy broke down the fundraising signals of roughly 5,000 U.S. startups on X. The sample covered about 60 attributes, and he ultimately kept the 15 strongest signals. The filtering criteria were strict: the seed or Series A round had to be led by a top-tier fund, and at least 25 companies had to have that attribute. As a result, Olympic medals, Thiel Fellows, and Forbes 30 Under 30 did not make the list because the sample size was too small. A note at the bottom of the chart says the data comes from Harmonic and covers 4,836 U.S. startups from 2021 to 2024.

The top item on the list is having worked at a top-tier fund, which raises the probability by 3.00x. This identity is broad and includes GPs, venture partners, EIRs, investors, scouts, scholars or fellowship program members, and even interns. Deedy's explanation is straightforward: you're close to the process, you know people, and you know how money gets invested. Someone in the comments asked whether this is simply because top-tier funds are already good at picking winners. Deedy responded that the sample only looks at seed and Series A rounds, and many companies did not yet have product-market data at the time, so it was impossible to pick people based only on later performance.

The second-strongest signal is coming from Tesla or SpaceX, at 1.83x. The signal for frontier labs is stronger, but the sample is too small. Caltech is the strongest university signal, while Waterloo is actually negative. Technion is very strong, but Israeli education overall is a negative signal. Former Stripe employees are a negative signal, and Deedy said he has not figured that out either. Being a former founder, even with an exit, is not as important as people imagine; part of the reason is that many exits are just small acquihires.

Among the negative signals, MBA is 0.60x, PhD is 0.75x, founders over 50 are 0.66x, and a fintech company background is 0.74x. A Technion or Tel Aviv U. background is positive, but Israeli education overall is negative. This combination looks contradictory, and Deedy also listed it as one of the interesting findings. Olympic medals, Thiel Fellows, and Forbes 30 Under 30 did not make the list, again because the sample was not large enough. To make the list, at least 25 companies had to have that attribute.

When Astasia Myers reposted it, she added a timing gap: the data range is 2021 to 2024, mostly before the AI wave. At that time, a PhD was a negative signal. Looking again now, the story has changed. Gaurang in the comments put it more precisely: the credential label has not changed, but the narrative around it has. A PhD used to be read as slow; now it is read as a moat. This may not apply to every industry, but at AI companies, the pricing of technical depth is indeed changing.

This chart measures who is more likely to get money from top-tier funds. It does not measure who builds good products. RAY in the comments put it bluntly: this is a chart about who gets let in, not about who builds something good. Dusan asked how to separate the top-tier fund signal from the fact that funds are already good at picking winners. AI Quanting also pointed out that the three largest effects correspond to the smallest samples—183, 147, and 132 founders—while variables with thousands of samples are closer to the middle. So the chart can show trends, but it should not be treated as an iron law. Mladen also reminded people that the data only covers U.S. startups; in Europe, CS degrees and the ex-Google line may be completely different.

The most worth pondering of these signals is time. From 2021 to 2024, both PhDs and MBAs were on the negative side. After the AI wave took off, the reading of a PhD shifted from slow to moat. The VC signal table has not changed; the market's pricing of technical depth has. As for which label will be revalued next, people in the comments have already started placing bets.

Sources: Deedy's X post https://x.com/deedydas/status/2116577618816577767 and Astasia Myers's repost https://x.com/AstasiaMyers/status/2116601596352299155.