Friends With(out) Benefits

You've stumbled on an amazing opportunity, or maybe you've been working tirelessly to bring it to life. But now you need help. Who are you going to choose to help build your idea into a viable business? It's a simple question without an easy answer: there are issues of expertise and availability, trust and goals, working styles. There's also a wealth of research suggesting that choosing the right partners makes a world of difference for both firms and individuals. And that's what makes a recent study of the venture capital industry by three Harvard researchers so interesting. Paul Gompers, Yuhai Xuan, and Vladimir Mukharlyamov assembled a dataset of 3,500 early stage investors and the 12,000 deals they collaborated on. The goal was to determine what factors made investors likely to work with one another, and how that affected their financial performance.

Their first big finding is that investors prefer partners with whom they have something in common: having worked together in the past made them 60% more likely to co-invest, and having the same alma mater or belonging to the same ethnic group each gave a 20% boost. These findings are in line with a general theory in social networks known as homophily, which refers to the fact that we generally like spending time with people who are like us. Homophily governs a lot about everything from how people find friends to how firms make alliance decisions, and it makes sense: having similar backgrounds and cultures allows for more effective communication, increases the likelihood of having common goals, and cultivates trust between partners.

But it's also dangerous. After all, as we've discussed in earlier posts, diversity is a big component of creativity and innovation. Homophily, by virtue of pairing similar partners together, limits access to diverse information and skills. And that's exactly what our Harvard team finds: the odds of a startup having a successful IPO are 22% lower if the investing pair are from the same university, and 18% lower if they share a past employer. Ethnicity has no impact on performance by itself, but when a pair of investors are from the same ethnic minority their startup is 25% less likely to succeed.

What this means is choosing partners is no simple task. On one hand, we need to find collaborators that we understand and trust. On the other, we need to find partners with complementary skills and diverse perspectives. Performing well, be it on a collaborative music project or a VC investment, means balancing these two factors. And that may take some creativity.

Check out the full paper here, or the brief Economist writeup that led me to it here.