A Diversity Report for Angels
In May of 2014, Google produced a report that covered the gender and ethnic backgrounds of their 50,000+ employees. This was the first of what would become a collection of annual diversity reports from many top tech companies.
The findings in these reports were … disappointing, to say the least. In fact, TechCrunch recently suggested that tech companies need a version of The Rooney Rule, which requires NFL teams to interview underrepresented minorities for open front office positions.
All of this focus on diversity at large tech companies is great, but what about the small ones? What do we know about how gender and ethnicity affect your likelihood to get funded, and how that varies across VC firms?
The answer, at least publicly, is not much. CrunchBase, NVCA and and Dow Jones VentureSource have joined together to launch the 2015 Venture Census, which aims to fill this void. According to TechCrunch, “the survey focuses on characteristics like gender, age, sexual orientation, educational and professional background and military experience.”
This is a great opportunity to get some hard data, and hopefully we will see some results in the near future, but is an exercise like this really necessary? Does anyone doubt that Danielle Morrill and her team at Mattermark have information around age/gender/ethnicity for a number of the companies they track? Particularly for the high-profile and venture-backed companies, I’d be shocked if they didn’t.
And this is what got me thinking about if and how someone could gather this information. You would need information on top investors, who they invested in, who had founded that company; basically a whole social network’s worth of connections — preferably with dates so you could see how trends changed over time.
A “List of Angels” if you will. An AngelList.
And that was the “a ha” moment.
So I did what any journalist COULD have done and went through all of the listed investments for the top-5 individuals on the list (Gil Penchina, Tim Ferriss, Jason Calacanis, Dave Morin, Naval Ravikant). I then broke out when each person invested in a company, as well as the gender and ethnicity of the founders.
I’ll have some nice graphs for you in a second, but first let’s cover how the worked.
Obviously there is a lot of opportunity to get this wrong, so I want to be clear that my work was more than just some quick approximation. I researched 968 founders across 436 different companies and put enough work into each one that I was either comfortable with what I marked down or excluded them from the analysis.
Not every founder has a profile picture on AngelList, so I clicked through to Twitter, Facebook, and LinkedIn profiles. If that didn’t provide any clarity I google’d the company name and founder name, hoping to find a piece of press with a headshot. If the pictures were unclear, I looked up the ethnic background of first and last names to see if I could get any hint. In all this probably took 30–40 hours of work, which led to my second “a ha” moment: oh, so THAT’S why someone trying to turn out 5 stories a day for tech site didn’t do this.
Genders were kept simple: Male and Female.
Ethnicities were: White, Black, Latin, Asian (which I classified as China, Korea, Japan), South Asian (Pakistan through Indonesia), and Middle Eastern.
For this analysis, I included Israelis in the “White” designation, which you could certainly argue is incorrect. However, at the end of the day Israelis (particularly the ones who showed up in this sample) tend to have a skin tone similar to the standard “White” archetype, and the idea here was to see how appearance may impact fundraising.
Also note that I did not distinguish between who was in the CEO role and who was “just” on the founding team. As many people have covered, angels fund teams. If you are on a founding team, sophisticated investors (which clearly applies to the 5 covered here) are looking at you when they make their decision.
With everything collected, I dumped the company/personal variables and angel connections into a local MySQL database, and got to work querying.
All of this data is presented in aggregate. The goal of this exercise was not to single out a “best” or “worst” investor for diversity, but just to see what trends may exist.
First, a look at the data by company.
If a company has a single non-white founder, they show up in the minority ranking, even if the rest of their team is 132 white guys. In this analysis, minority is only representative of ethnicity.
Raw numbers aren’t always the best, so I also looked at the percentage of funded companies with at least 1 founder in any of these groups.
Next I wanted to look at the breakdown by founder.
The idea here was to look for any difference in the rate at which companies with at least one underrepresented founder are being funded, and the rate at which people from those groups are joining the startup ecosystem.
And again, a second look at the data by percentage of founders. I’ve kept the scale the same as the “company” graph to help show differences.
It seems clear that the focus on female founders over the last 18 months has had an impact. The percentage of companies with at least 1 female founder that received funding from this subset of investors more than tripled from 2013 (7.8%) to 2014 (28.0%) and remains over 20% at this point in 2015. However, when we look at the numbers by founder rather than company, the percentage of female founders has only barely crossed over 10%.
Unfortunately we don’t see much of any growth in representation for minority founders, whose participation percentage has actually fallen on both a company and individual basis since 2012. This is especially true with black and latin founders who show up in less than 3% of funded companies so far in 2015 and make up just 1.1% of founders.
You could spend a lifetime trying to explain why this happens and attempting to assign blame. That’s not the point of this post.
I truly don’t believe that any of the 5 people listed would or has turned founders away solely because of their gender or ethnicity.
However, I do think that getting the numbers out there can help investors to see some of their own patterns and maybe act on finding ways to change them. Maybe they look a little harder for investments that aren’t so easy to spot. Maybe they give one more look to a company they initially wanted to pass on.
I also think that initiatives like Tristan Walker’s and Laura Weidman Powers’s Code 2040 and Jen Bonnett’s Startup Chicks have the opportunity to be supremely impactful in the other direction; driving more people from underrepresented groups into tech startups. These new voices bring with them a lifetime of different experiences and understanding around the problems they want to solve and the customers they want to build for.
And everyone wins when that happens.