I was recently sitting in a room at a pre-seed summit, surrounded by some of the best early-stage investors in the U.S.
The conversation drifted to a question I keep coming back to: how do you actually make a decision when the signals are mixed? Over and over, the best investors in the room gave the same answer: their best investments were made against the grain.
The best early-stage investment decisions happen before the data agrees.
People talk about VC like it's a data problem. And at later stages, it increasingly is—you have revenue, retention, CAC/LTV ratios, comparable exits. But at pre-seed, you're often looking at a founder, a hypothesis, and a whiteboard's worth of conviction. The numbers don't exist yet. You're often not underwriting a business. You're betting on a person and a belief.
For example, Masayoshi Son invested in Alibaba after a single meeting with Jack Ma because of the look he saw in Jack’s eyes. I’m not kidding. It went on to become one of the most valuable investments in VC history.
Most great early-stage investments don't look great early. They look weird, niche, too early, or just confusing. This is the tension of early-stage investing: by the time something looks like a good investment, it often is too late.
The further upstream you invest, the more you are operating in a low-data environment—by definition. Over time, you earn pattern recognition, which is helpful, but it’s more data. Data can kill the best early-stage bets, so the ability to review opportunities without bias or data is when magic can happen.
I've found that the investors who consistently find the best companies early aren't necessarily the most analytical people in the room—they're the most curious.
The more you know—across industries and disciplines—the more surface area you have for something a founder is building to genuinely resonate. The curious and the creative lifelong learners make the best investors.
At Redbud, our buy box is almost entirely built around human criteria as we believe the greatest driver of venture returns is resilience. That's a deliberate philosophical statement about where alpha comes from at pre-seed.
We look for resilience, integrity, high velocity of execution, paranoia, a little bit of crazy, unique distribution, and a real secret about an industry. Network effects are a nice-to-have, but they're the one structural variable on the list — everything else is about the person. But deep down, the entrepreneur has a secret to build something special, so is it really that risky?
I’ve been an early employee and co-founder of a few startups, and my partners built EquipmentShare. As a VC, having an entrepreneurial background is incredibly helpful. The more you understand what it feels like to build, to fail, to pivot, to fight for something you believe in against long odds — the better you are at recognizing it in someone else across a table.
The problem with only using data is this: data describes the world as it was. But at the frontier (where the best early-stage bets live), data doesn't exist yet. The VCs who win consistently are the ones who allow conviction to override data.
They can sit across from a founder building something strange and unpopular and ask themselves: not 'does the market agree with this yet?' but 'is this person seeing something real that the rest of us aren't?'
By the time it’s obvious, it’s over.
