While trawling through HackerNews on a recent afternoon, I stumbled upon an interesting dataset. Published by FirmAI on Kaggle, the data contained information on over five hundred startup founders, with nicely tagged and cleaned data detailing their past work and educational experiences.
With the recent rise in “unicorn” startups elevating many of their founders and chief executives to near-mythic levels in the startup/venture capital world, I thought it might be fun to use this data to de-mystify who these founders really are!
Notes on the Data
- There are 574 distinct founders in the dataset - some of which I recognized at first sight (or their company), many of which I could not
- As such, while it’s tough to know whether this data is truly representative of Silicon Valley, it seems to be relatively broad-based
- Survivorship bias is definitely in play here - by focusing our attention on founders and companies large enough to be notable, we miss a lot about the demographics of founders and companies which didn’t quite make it. As such, we should all understand that this data is going to be biased in the direction of “successful” startups, and accordingly the kind of founders which are more likely to be financially successful
- To tag founders as male or female, I manually reviewed and annotated each founder
- While the
startup_foundersdataset seemed very accurate upon cross-reference online, the
startup_valuationsdataset showed significantly different numbers than were reported on Crunchbase. As such, we don’t rely on any of that data to be part of this analysis!
- Even though I expected the ranks of founders to skew heavily male, I was shocked at how heavily male they actually ended up being - well over 90% of the founders in our sample were men
- Almost 50% of founders had previous startup experience - not shocking considering how exposure to investors and funding networks can grease the wheels for future investment
- Nearly half of founders have received graduate education, often in business (scroll down for more on that!)
Despite my high hopes for uncovering some nifty correlations in the data, we don’t see many strong correlations when comparing all variables against each other. In my opinion, the two most meaningful takeaways from this data are that:
- Stanford and Berkeley graduates have a relatively high likelihood to have worked at Google in the past, and
- Those that have worked at Google in the past are relatively likely to have been a Product Manger, Head, or VP in their previous work experience as well
Given that the connection between Stanford/Berkeley and PM/Head/VP is also relatively high, it seems reasonable to infer that there is a “pathway” amongst this cohort of people into founderhood. Interestingly, the Ivy League does not see this same effect - there appears to instead be a funnel into Consulting as a pre-startup pathway.
- Roughly 88% of founders attained an undergraduate education at a college or university
- Amongst that group, computer science stands out as by far the most popular field of study
- At the undergraduate level, relatively few undergraduate degrees were in business (roughly ~10% of all founders)
- This is largely due to the long tail of other disciplines that founders were interested in, from Electrical Engineering to International Relations to Design
- However, the graduate level of education sees a significant amount of concentration in business, with roughly ~60% of graduate degrees being MBAs
- With roughly 44% of founders attaining graduate education, over a quarter of founders are therefore MBA holders