Oxford, Cambridge, Imperial: who do our founders build with?

I've been digging into some data on founding teams with an Oxford-affiliated founder, and a few patterns stood out—some expected, others genuinely surprising.

Using Tracxn (a venture capital database), I looked at which other university and company affiliations appear alongside Oxford alumni on startup founding teams.

I also compared these trends with Cambridge and Imperial to see how Oxford stacks up.

It’s all observational—correlation, not causation—but it offers some interesting signals about the founder pathways shaping the UK’s venture ecosystem.

Who do Oxford founders build with?

Turns out, if there’s an Oxford grad in the founding team, the most likely additional university affiliations are Harvard (9.7%), Cambridge (8.1%), and Stanford (5.2%). This isn’t shocking—Oxford and Harvard have deep academic ties, and the Golden Triangle (Oxford, Cambridge, and London) naturally sees a lot of cross-pollination. Imperial and LSE also make notable appearances at 4.8% and 4.2%, respectively.

Cambridge founders, by contrast, are slightly more likely to be building alongside Imperial or UCL alumni—5.1% of Cambridge-affiliated teams also include an Imperial founder, compared to 4.8% for Oxford. Imperial, as expected, has a strong connection with both Cambridge and UCL, with around 10% of Imperial-affiliated teams also including an Oxford, Cambridge, or UCL grad.

This tells me two things. First, Oxford founders are more likely to co-found with someone from another elite international institution (Harvard, Stanford) rather than sticking strictly within the UK network. Second, Cambridge’s startup ecosystem seems a little more intertwined with the London-based tech scene, which aligns with its strength in engineering-heavy startups.

BCG surprise

Now for something unexpected: Oxford founders are disproportionately likely to have a BCG alum on the team. In fact, 8.7% of startups with an Oxford founder also include someone who previously worked at Boston Consulting Group. That’s nearly double the McKinsey representation (4.8%) and higher than Goldman Sachs (4.4%).

If I had to guess why, it’s probably a mix of things. BCG has a strong footprint in London and has been particularly active in deep tech and climate tech strategy work—two areas where Oxford founders tend to launch startups. It’s also possible that the network effect from Oxford’s PPE and business degrees feeds into this pipeline more than other consulting firms. (Will look into this).

Meanwhile, Oxford-founded startups are twice as likely to have a co-founder from a big bank (Goldman Sachs, Morgan Stanley) than their Cambridge or Imperial counterparts. Cambridge founders, in contrast, are more likely (by about 20%) to have a FAANG or big tech alum in their founding team - although the numbers are tiny so this might be quite a nominal difference.

The takeaway? While Cambridge founders skew more towards engineering-heavy teams with big tech experience, Oxford founders seem to bring more finance and strategy expertise into the mix. Imperial sits somewhere in between.

Why this matters?

These patterns reinforce my broader interest in the Golden Triangle as the UK’s venture flywheel. The UK government’s renewed push for the Oxford-Cambridge Arc only makes this conversation more relevant. Understanding where founders come from—and who they’re likely to co-found with—helps identify how to strengthen the startup ecosystem and plug in the various hubs and spokes of the flywheel.

Is this interesting? I think I'll go deeper over the coming months pulling more granular data on industry, valuations, and capital raised. I also want to explore whether AI tools can help map ‘founder pathways’—what common stepping stones (big tech jobs, second-time founders, accelerator programs) shape the most successful Oxford startups. Will be looking to scrape some data and do some (pretty basic) AI tooling to see what I can learn about the UK's venture flywheel.

If you’ve got thoughts on this—or want to collaborate on the next round of analysis—drop me a line, scrappy/rough calcs here. Calcs are WIP.