When investor Aileen Lee started her firm Cowboy Ventures in 2012, she saw it as a “clean slate” — a chance to focus on different companies from those she had over the past decade at Kleiner Perkins Caufield & Byers, and to check if common assumptions in venture capital were supported by the numbers.
There was just one problem: She couldn’t find any numbers.
“The lore of venture capital is, you work really hard with a diverse portfolio of companies,” Lee said on the latest Recode Decode, hosted by Kara Swisher. “You might have 20, 30 companies in a fund, and at the end of the day, it’s like two that matter. Like, can I see the data?”
Lacking, at the time, any tools like Crunchbase, Lee did the homework herself, aggregating news articles and calling entrepreneurs directly to understand who started companies, where they came from, how much money they raised, how their companies were valued, whether they pivoted over time and whether they succeeded.
That research confirmed that only a tiny fraction of startups, which she dubbed “unicorns,” made it huge. Less than one percent of companies would be valued by public or private markets at over $1 billion within ten years of their founding.
“It may feel like it’s an overused term, but that is still what our industry is about,” Lee said. “From an investor perspective, particularly for traditional venture funds that are $600 million, the way the math works is you have to be investing in unicorns. If you have a billion dollar fund, you have to own a certain amount of a $10 billion company to be able to be doing a good job.”
However, other common VC assumptions were completely refuted by the data. Most notably, she found that startup success was correlated with founders who already had work experience.
“At the time, in 2013, there was this ‘Drop out of college! Start a company! You’re a loser if you haven’t started a company by the time you’re 24’ kind of a thing,” she said on the new podcast. “These people started their companies in their 30s, and they had work experience beforehand, mostly at tech companies.”
“I would be having meetings with founders coming in, being like, ‘Hey, I’m really sorry, I’m 35, I know I’m really old, but I hope you’ll still meet with me,’” she added.
This article originally appeared on Recode.net.