A tale of two investors

It’s time for venture investors to dig deeper on diversity, starting with the right metrics.

Ellen K. Pao
Apr 25, 2019 · 6 min read

True diversity requires metrics and accountability. It’s more than intent, it’s more than optics, and it’s a lot more than a superficial measurement.

Venture capital firms are starting to understand and share diversity numbers for portfolio founders. Most, like USV and Brooklyn Bridge Ventures, are calculating metrics at the startup level, where they count the percentage of startups with at least one founder in a particular category. The Limited Partners that invest in VC firms haven’t required metrics, but they should; industry numbers show that most firms — even women-led ones—still invest mostly in homogeneous founding teams; diversity metrics would show what firms are actually making tech more diverse and equitable.

But while this metric is helpful, it doesn’t tell the whole story. Measuring diversity at the startup level inflates progress since most startups have more than one founder — which prevents us from seeing the problems and doing the work to fix them. The intention may be good, but it has an insidious impact: Its inflated values bring kudos to VCs for not-so-diverse portfolios. And it implies they and others don’t have to change what they’re doing — much less transform it.

Think about it.

Would we measure the diversity of a college faculty by counting only the number of departments with at least one woman faculty member? Or one professor of color?


What if we reverse the numbers and look at the percentage of startups with only male founders? What if we looked at percentage of startups with at least one male founder?

In one case the metric shifts from 33% of startups with least one female founder to 95% with at least one male founder.

It starts sounding a lot less diverse — and a lot more problematic.

It sounds like a whole different investor. And it points to a need for change.

The comparison shows how percentages differ significantly, especially for female founders (26% under the startup method compared to 14% using founders).

Measuring dollars is critical because of what we already know.

Because investors, especially VC firms, allocate time and resources based on the size of the investment, lower dollars usually means less help and less support in addition to less runway.

We have so much work to do, and that means we need make sure the metrics are right.

Divide dollars invested in a startup by the number of founders before allocating by gender and race of each founder.

  1. Measure significance. Calculate % of $s to founders of underrepresented groups divided by total $s invested. Take dollars invested in women and Black and Latinx founders as calculated above (don’t double count) and divide by the total dollars invested in the same time period.

Here’s what it looks like using my own portfolio of 24 angel investments.[1]

When I measure who my money goes to by demographic, I see helpful differences: I’m investing fewer dollars in entrepreneurs of underrepresented race/ethnic groups and in nonbinary entrepreneurs than I would want (10%, 10%, 5%). The differences are not as huge, but at least now I know where to pay attention to figure out why and how I got here. I have not invested in any Pacific Islander or Native American founders that I know of, and that’s a gap I want to fill.

When I consider intersectionality, I find I have invested in 12 women founders of color (23% of founders), including 5 Black or Latinx women founders (10% of founders). I also looked at the demographics of the CEOs, where power in a startup is concentrated. I invested in 14 women CEOs (58% of the total number of CEOs), including 9 women of color CEOs (38%) with 5 (21%) from underrepresented racial/ethnic groups.

Now I know where I need to be more thoughtful and careful, and I can proactively and meaningfully address my biases. The simple metric tracking percent of startups fails at both. Over time these metrics will evolve as our understanding evolves. I’d love to add a metric of how much equity founders get by demographic, following on the work of #ANGELS on #TheGapTable. (I’m also open to suggestions for other metrics to consider or metric data from other investors; please include your ideas in the comments.)

Now that it’s all laid out here, the question for investors and the limited partners that fund them is:

When you look in the mirror, do you want to see the investor you actually are or only the investor you want people to think you are? Do you want true diversity and change or performative diversity and stagnation?

We want to gather these metrics for benchmarking and if you are willing to share, either publicly or anonymously, at the firm-level or for aggregation, please reach out.



[1] I categorized founders and CEOs based on how they have described themselves to me or publicly. I don’t assume race or ethnicity or gender other than what I have been told. I may miss identities using this process, but I worked conservatively and didn’t make assumptions. Also I did not include my first angel investment in reddit, because when I invested both founders had left and not yet returned. I did not include my latest investment, which I made after I drafted this post.

[2] For comparison, this last chart shows how demographic metrics vary by startups, by founders, and by dollars. Also, for completeness, flipping the startup metric, 60% of startups I’ve funded have at least one male founder.

Project Include

Give everyone a fair chance to succeed in tech.