Two weeks ago, I published a new report for the Center for American Entrepreneurship, titled The Ascent of Women-Founded Venture-Backed Startups in the United States. I followed-up with a summary on this blog last week.
The study analyzed venture capital first financings in the United States between 2005 and 2017, segmented by the gender composition of founding teams. It established two key trends. First, the number of “women-founded” startups entering the venture-backed pipeline each year—while unquestionably low compared with women’s representation in adjacent areas of the economy—is improving steadily over time. Second, once women-founded companies raise a first round of venture capital, they achieve similar outcomes as non-women-founded startups—such as the ability to raise follow-on rounds of capital or achieve an exit (IPO or acquisition). There is a lot more in the report and I encourage anyone who is interested to go and read it.
One criticism of the report is my definition of “women-founded”. For reasons I explain in detail in the report’s methodology, I chose “women-founded” to indicate a company that has at least one verified female founder. That means it includes startups with all-women founding teams and teams with both women and men (coincidentally, it also means that I assume that companies with missing founder information had no women founders—more on that in a second). A key reason for not separating these groups was needing a bigger pool of companies to draw from in order to credibly track outcomes over time—and there just weren’t enough of them in the mid-to-late 2000’s to do that. There were tradeoffs.
However, that does not prevent me from more narrowly segmenting these groups here and demonstrating first financing trends only across the four types of founding teams in the dataset—women only, men only, mixed gender, and missing gender. To begin, the first chart here displays the raw numbers of annual first financings for startups falling into each of those four founder-gender categories.
A few insights stand out. First, male-only companies dominate the pipeline for venture-backed startups in the United States and helped drive the massive expansion between 2009 and 2014, before tailing off sharply since then. Second, the number of both women-only and mixed-gender teams rose over the period—particularly after 2009—and startups with no information on founder gender declined over time (presumably as the data vendor, PitchBook, improved its collection methods).
Since it’s difficult to visualize the latter set of data in the chart because of the men-only figures, here we’ll exclude the men-only founded companies and present the same data—plotting the share distribution across the three other founder-gender types.
And we can see here, the number of women-only and mixed-gender founded startups increase dramatically over the period—expanding from 3% to 7% and from 4% to 15%, respectively. The chart also shows how much data collection methods have improved over time. Nearly one-third of companies in 2005 have no founder-gender information—a number that fell to 7% by 2016 before oddly spiking to 12% in 2017 (though I suspect this will fall again as more data are incorporated into this dynamic database).
Next, let’s look at growth in first financings, but here I want to bring male-only founded startups back into the fold. I also want to get rid of the “missing founder” category but I don’t want to toss out those companies altogether. So, I’m going to make a fairly non-controversial assumption and say that the founder-gender distribution within the missing category in a given year is identical to the founder-gender distribution within the “known” category. This allows me to maintain levels and growth, but assigns the missing founder firms to the remaining categories propionate to their “known” share.
With this adjustment in place, comparing back to the previous chart, this pushes the first financing shares of women-only and mixed-gender teams up a few percentage points—in the case of the former, from 4% in 2005 to 8% in 2017, and for the latter, 6% and 17%. When combined, this set of “women-founded” companies increased from 10% of startups in 2005 to 25% in 2017. Compare that to the “women-founded” percentages in the report—which did not make the adjustment for missing data—at 7% and 21% respectively. If we treat it as a range, we can say that somewhere between one-fifth and one-quarter of new venture-backed companies in 2017 had at least one female founder.
To get at growth, the chart here shows the percentage change in annual first financings compared with 2005 (the base year) for each of the three founder-gender groups—women-only, mixed-gender, and men-only. Each point on a line tells you the percentage difference between first financings in that year versus in 2005. Over the entire period, startups with mixed-gender founding teams grew the most in percentage terms, followed by women-only founded companies. Men-only founded companies are well behind. So, while men-only founded startups still dominate overall levels of first financings today, women-only and mixed-gender founding teams are disproportionately driving growth.
Finally, to get a closer inspection of changes underlying the longer-term trend, the table below shows changes in first financing activity by founder-gender over various time periods. The figures have been computed as compound annual growth rates in order to compare changes occurring over differing time periods on even footing. As the data here show, mixed-gender and women-only startups have been the most resilient in terms of growth rates throughout the period. Women-only teams have exhibited the strongest growth the latest three years as the rest of the market went into decline.
As I wrote about in the report, the headline numbers for women in venture capital and as founders of venture-backed companies are bleak. In 2017, just 16 percent of the nearly $83 billion invested in U.S. venture-backed startups went to companies with at least one female founder, and only 2.5 percent went to startups with all-female founders. Also, a measly 9 percent of general partners at leading U.S. venture capital firms are women. All is definitely not well.
However, as this analysis demonstrates, there are positive signs underlying these more discouraging headline numbers. Considerably progress is being made in diversifying the venture-backed startup founder base, even if there is still a long way to go. I’ll have a few more posts that go into other facets of the data in the coming days and weeks.