The result, they hoped, would be not only a more data-driven model for finding portfolio companies but also a more diverse group of founders and companies.
Eight years later, Ulu Ventures has delivered on both fronts. Of the companies in Ulu’s current portfolio, 30 percent have women CEOs, while 30 percent are led by minority CEOs and 10 percent have Latino and African-American co-founders.
We recently sat with Miriam and Clint to learn more about analytics, decision-making and their best advice for other early-stage investors looking to take advantage of data.
First, why did you want to bring more analytics into early-stage investing?
Clint: In addition to being an entrepreneur myself, I also did my Ph.D. at Stanford, where I focused on studying decision-making under uncertainty. All the companies I founded centered around tools to help organizations make better decisions. When you look at big industries such as pharmaceuticals or oil and gas, they have best practices for making decisions. Then you look at venture capital where they rely on “pattern matching” or insist, “you just have to have the magic” — which is basically synonymous with huge cognitive bias.
Add in the fact that so few investments get made, using a really small number of a data points and a terrible environment for learning, and it’s clear why some VC investors have these stubborn built-in biases which lead to fewer investments in minorities and women who don’t match the pattern. The way to counteract this is with decision analysis or a systematic structure for making investment decisions. This allows us to focus our attention on the key drivers of value.
What does that decision framework look like in practice?
Clint: First, we do a market-mapping session with entrepreneurs. We sit down and create a graphical picture of their target market, competition, how it will change over time, etc. — essentially, we quantify everything. We also have a decision model we use internally that accounts for things such as the company’s market share, revenue potential, team risk and more. We end up with a weighted multiple on the investment capital, which we use to help determine whether and how much to invest.
How has that formula worked in terms of predicting successful investments?
Clint: Yes, for example, we had one company that had attracted a lot of other investors, but we ran our analysis and it showed only a 3x potential return on the invested capital. We thought that maybe we were missing something, because there were all these important people involved — and we ended up losing money. In another instance, we had a company go through the analysis and it showed a 17x potential return. We thought that it had to be too high and invested half of our normal amount. Now we’re kicking ourselves at the fact that we didn’t invest more.
Miriam: The failure rate of the entrepreneurs that we went through the process with in Fund 1 was 10 percent. The failure rate for those that didn’t was 50 percent. In fact, some of the biggest mistakes we made in the first fund transpired when we didn’t listen to our analysis.
How does the framework affect the diversity in your portfolio?
Miriam: We end up investing in a diverse array of companies because we do the same kind of analysis, regardless of the type of business. But beyond the decision framework, I think that more diverse funds tend to attract more diverse entrepreneurs. If you see a diverse team and you happen to be a diverse team, then you are more apt to feel like you’ll get a fair shake.
We’re doing more in diverse communities and communities of color. And because of that, we have more diverse networks. Frankly, our pipeline is actually more representative of the industry as a whole. Then you take this pipeline back to our process, and we have these criteria that measure the same things for every company — which make it hard to fall back into any inherent biases.
What advice do you have for other investors in terms of employing more data in their decision-making?
Clint: Our investment thesis is about smart data — not big data. Investors don’t need a lot of data. In fact, the only value in data is when it helps you make a better decision. The key is structuring your thinking in a disciplined way and creating a framework that accounts for uncertainties. You don’t need a lot of data, but you do need some.
As a final thought, at Ulu you also emphasize the importance of softer skills in creating a company — something less tangible than data. Why is that important?
Miriam: Qualities such as ethics and empathy can often be overlooked and underestimated in Silicon Valley, where many people tend to think that sheer intellectual horsepower is the key to creating value. For a lot of these companies, the shelf life is so short.
It’s people’s very life energy that is being used to create these startups, and you have to be able to set up an organization that recognizes that. It’s the difference between something having real meaning and purpose, and something that’s just an ego trip.
Hana Yang First Republic Bank interviewed Miriam Rivera and Clint Korver Ulu Ventures
(This column first appeared in May 2018 on the First Republic Bank site)