Just thought this was interesting. The below is taken from a facebook post (page: Boardroom; I’m on my phone and not going to search for the source doc right now). My own random observations: 15 out of the top 20 are American, with two Israeli universities in the top 10 and 2 Canadian and IIT (India) also in the top 20. But no European or Chinese universities…
UC Berkeley is still officially the world’s top startup factory.
PitchBook’s latest university rankings once again place Berkeley at No. 1 among schools producing venture-backed founders, with Stanford, Harvard, the University of Pennsylvania, and MIT rounding out the top five.
The rankings, based on data from more than 173,000 VC-backed founders, also show that the startup landscape is becoming more competitive globally. The University of Texas moved ahead of Michigan this year, while Canada’s University of Toronto and University of Waterloo both landed in the top 20.
(via Pitch Book Data)
It’s not surprising to me, considering that it’s only undergraduate and listing total, rather than per capita. UCB has 33k undergraduates. Stanford and Harvard 7k undergraduates.
Clearly it’s more than just numbers, or the rankings of at least some of the other public colleges on the list would be higher, and we would be seeing Chinese universities etc there too.
Berkeley is great school located in Silicon Valley, with a large number of students majoring in fields associated with start ups. I’m not surprised that Berkeley is higher than other public colleges. My point was more about why Berkeley was “officially the world’s top startup factory” as listed in the title, rather than Stanford, Harvard, or similar.
I was more interested in how despite the fact that so much is made of Chinese universities catching up to US ones /how fast they are rising in global rankings, etc, they don’t make an appearance. Obviously part of that might be the institutional system in China, given these are VC funded startups, but I also wonder if it points to the US still being the clear innovative hub globally. Definitely various things to take away from this. Location is some advantage for sure but clearly not definitive either - given the appearance of colleges like Cornell and UIUC on the list, but of course most startups happen after students have graduated and presumably gravitated towards the tech (and financing of tech) hubs.
Sorry for the plug, but when adjusted for school size…
Where Babson Ranks
#1 Globally Per Capita: When adjusted for school size—measuring the number of entrepreneurs created per student over the last 5 years—Babson College ranks as the No. 1 university in the world, outperforming institutions like Stanford, MIT, and Harvard. [1]
Pitchbook’s methodology isn’t picking up much of the activity in China for a few reasons. Pitchbook is tracking only commercial venture capital, whereas in China a significant proportion of funding comes from government sponsored vehicles.
Pitchbook also relies on regulatory filings and English language press releases. Pitchbook is also not picking up capital funds raised by Chinese companies in Renminbi (there’s been a huge shift in China over the last decade for entrepreneurs/startups to raise capital in RMB).
Point being, Pitchbook’s methodology is not picking up significant proportions of startup activity in China. That does make me wonder about the credibility of this entire database, but that’s me…I know Pitchbook has cred in the real world. Perhaps an example of the ‘best’ available still not being very good/accurate/comprehensive.
I also agree with @heavyweightcollege that it seems these data should be adjusted for size of student body.
Thanks for the background on China /methodology notes - so there is a lot missing beyond just the VC factor. I wonder if that also means Europe is being undercounted (both language and perhaps funding type.)
The linked references do not appear to support the comment. Babson didn’t appear in Pitchbook’s top 100 for undergraduates. Prior to dropping of the list, Pitchbook was ranked 96th with 146 founders. If I assume Babson is ranked 101st today with 188, then the per capita totals are as follows.
The numerator is the PitchBook number of founders as referenced in original post of this thread. The denominator is number of undergraduate students. There is some methodology information on PitchBook’s website at https://pitchbook.com/news/articles/pitchbook-university-rankings .
Does not add up for me, so I reached out. Will share the response I get, if I get one. I have worked with them over the years. They have become more of a revenue generating organization, so curious to hear their explanation.
Babson UG teaches more of a bootstrapping culture, rather than venture capital, so maybe that is why. Also, believe this data comes from more than a single year of grads (need to confirm). So if 10 years, for example, that’s 28k vs. 330k, right? Please correct me if I am wrong. Or should that # be adjusted for the graduating classes, only?
Could be. Only looking at commercial VC funding as Pitchbook does is limiting…I’m seeing various estimates from 0.5% to 5% of start-ups ever get VC funding. People bootstrap, raise money via friends and family, and/or are quickly profitable and never need VC money. VC money is also skewed towards tech…sources say greater than 50%, so that partially explains the presence of relatively more tech-y schools on this list.
Obviously VC data is what Pitchbook has access to, so makes sense that’s the data they would use. It’s a nice little project for the interns
Yes, their methodology says these rankings are based on commercial VC funded activity between Jan 1 2014 and Sept 1 2025.
Another methodology factor that may skew results is Pitchbook doesn’t consider total capital raised per founder, just raw number of founders who have raised any money thru VC firms. So, if one analyzed VC dollars raised per founder, the results could look different. Here’s a blurb from a few years ago looking at Claremont Colleges’ VC $:
If 2024 Was Adjusted for Size, Would Babson Be #1?
Yes, absolutely. If you normalize the 2024 PitchBook data by school size, Babson College would be the undisputed #1 undergraduate startup ecosystem in the world. [1]
The raw math from 2024 shows why:
The Competition (#1 UC Berkeley): Had 1,804 founders, but a massive student body of roughly 33,000 undergraduates. That means 5.4% of the student scale produced a venture-backed founder. [1]
Babson College: Produced 230 raw undergraduate founders out of a microscopic student body of roughly 3,100 undergraduates. That means an astonishing 7.4% of its student scale produced a venture-backed founder.
No other school on PitchBook’s Top 100—including Stanford, MIT, or Harvard—reaches a 7.4% per-capita founder density. [1]
What Happened to Babson in 2025 and 2026?
Babson actually dropped completely off PitchBook’s public Top 100 undergraduate list in 2025 and 2026.
While that sounds alarming, it is a byproduct of how PitchBook changed its tracking, combined with the math of a small school:
The 2025/2026 Data Shift: In the 2025 PitchBook Global Rankings (which carry into 2026), the threshold to make the Top 100 undergraduate list skyrocketed. The #50 spot (University of Oxford) required a minimum of 332 raw founders. [1, 2]
The Small School Penalty: Because Babson only graduates about 800 total students a year, it physically cannot keep pace when PitchBook’s raw volume threshold jumps from 200+ up to 330+.
What Might Have Happened Behind the Scenes?
Because PitchBook hides schools below the Top 100, we don’t see Babson’s exact 2025/2026 undergraduate numbers. However, based on venture market trends, two things likely happened:
The Venture Capital Downturn: The venture capital market experienced a massive pullback in 2024 and 2025. When VC funding dries up, investors heavily favor founders with several years of corporate tech experience (favoring larger engineering schools like Waterloo or UIUC) over younger, newly graduated founders. [1]
The “Accumulation” Gap: PitchBook uses a rolling 10-year tracking window. If Babson had a massive, highly successful cluster of graduates from 2013–2014 who raised money, those founders aged out of PitchBook’s tracking window by 2025/2026. Because Babson’s total alumni pool is small, losing just 20 or 30 historical founders drops them off the volume-heavy leaderboard entirely. [1, 2]
This is one of the problems with using AI. The response sounds like it is an expert, but if you look in to the specifics, there are numerous errors. Some of the problems with AI’s response include.
It’s a 10 year total, not 1 year total; so AI’s percentages for all colleges are 10x too high.
Number of undergraduate founders is wrong. Babson was not among top 100. #100 was 189, so Babson was under 189, not 230. Perhaps AI made a bad guess about number of founders since there was no reference in table to look up. However, 189 vs 230 is still rounds to 0.7% (not 7%).
AI did not compute the totals for MIT, Harvard, Stanford, and other colleges which were all higher than Babson with errored method discussed above. Note that my totals for Babson and Berkeley from earlier post match AI, when rounded to nearest percent. I listed Berkeley was as 5% and Babson was listed as 7%, like the AI. However, I listed MIT as 25%, Stanford as 21%, and Harvard as 19%; and AI does not include these colleges (need to divide by 10 to show % founders).
Reasons for the problematic AI response include the leading phrasing, " If 2024 Was Adjusted for Size, Would Babson Be #1?". And it’s time/computation intensive to compute the totals for all 100 colleges, so AI only compared Babson and Berkeley. If you asked a followup prompt asking AI to show the colleges in a table that lists the totals for each college and “think slow”, the AI response would be very different.
More likely contributing factors to the discrepancy include the list only considering undergrads and different definitions of “founder”.
I’d generally agree we need to be careful about seeing this as a value-added measure, even in fact if we normed it per capita.
The US remains a fantastic environment for getting startup funding. And then the US (and sometimes International) kids most likely to go down that path disproportionately choose certain US undergrads, at least when they can.
So a college like that being a step somewhere in the middle of that process doesn’t necessarily mean a lot of unique value is being added at that step. Not that getting a college education is completely unimportant (although there are some interesting examples of people quitting college or only going to college later). But the degree to which these colleges are actually adding more value than others, instead of just being colleges such kids tend to choose, is very hard to sort out.
None of which implies they are BAD choices for such kids. I think this observation is most relevant when, say, there is a major cost barrier. Like, if your flagship isn’t in that graphic, and your family isn’t very wealthy, should you actually pay OOS tuition for Cal instead? I think often that is not really necessary, and therefore not really justifiable (not on these grounds, at least).
This makes it sound like a bug, rather than a feature. If you are looking at something like this, surely you don’t want the numbers distorted by what might be one or two successful years - aren’t you looking for consistency/continuity?
It is not clear to me btw if the example above actually happened or if it’s just the Ai making up an example.
Agree, but it probably becomes self-reinforcing at some stage.
I think I saw on cc the other day someone mentioning a study that showed the colleges that kids got into (not necessarily the ones they went to) was the best predictor of success, in other words it is the student not the college. Plus the intangibles…sure, Silicon Valley is probably the best place to be if you are looking for people to fund your tech startup. But what is the concrete difference in outcomes if you went to college there vs if you ended up there after you graduated? Is this measurable even?
I’m assuming from the methodology as described above that there is also no follow up tracking to see the number of these startups that were ultimately successful? I have a friend who works for a deep-pocketed company that is also involved in this. He was saying a while back something like - 20 start-ups come to you with ideas. You give some money to all of the ones that sound reasonable, in something like a spray-and-pray approach, even though you know most of them will likely fail, because you only need one of them to actually pay off. So all 20 would be in the pitchbook data I assume,but then what?
And to kind of combine your thoughts, if you aren’t living in, say the Bay Area already, maybe you should move there after college if you are serious about hustling for startup money. But moving has out of pocket and often opportunity costs. If there is only like a 5% chance (or whatever) of your startup even succeeding, maybe you decide against it. But if instead you are living there already, the cost of taking a shot on it is probably a lot lower.
But I do know people who went to Midwest publics and such who planned from the start to move to coastal cities eventually, and some of them succeeded. It isn’t impossible, it is just understandably less frequent.
