Investors call it pattern matching, and researchers call it bias. The last deal that worked shapes the next one, founders who resemble you feel safer, and a strong pitch can outweigh weak unit economics. In the largest survey of VCs the team ranked as the number one decision factor, which also makes it the most subjective one.
Experience does not remove these effects. If anything it makes them faster. A structured second read helps because it has no favourite: it applies the same criteria in the same structure, whoever pitched and whoever reads. The partner still decides, with fewer blind spots.
"She reminds me of..."
Sit in enough investment meetings and you will hear some version of this sentence: "She reminds me of the founder of that company we backed in 2019." It is usually meant as praise, and it is usually called pattern matching. Investors are proud of it, with some reason, because recognising what a good team looks like is a real skill that takes years to develop.
The difficulty is that psychologists have another name for judging a new case by its resemblance to an old one, and the name is bias.
What the research says
In the largest survey of how venture capitalists decide, covering 885 investors, the management team ranked as the most important factor, ahead of the product, the market and the business model. That makes sense. It also means the factor that matters most is the one that is hardest to measure and easiest to judge by feel.
Three effects show up again and again in research on decisions like these. Recency means the last deal that worked, or the last one that failed, colours how you read the next one. Affinity means founders who resemble you, in background or manner, feel like a safer bet. And a confident, charismatic pitch can carry weak unit economics further than it should.
None of these effects goes away with experience. If anything, experience makes them faster, because a seasoned investor reaches a conclusion sooner and with more confidence.
What actually helps
Telling people to be less biased achieves very little, as most investors know. Structure achieves more: deciding in advance what you will look at, looking at the same things for every company, and writing the assessment down before the discussion starts, so that the loudest voice in the room cannot rewrite it.
This is where a second read is useful, and it matters that it comes from somewhere with no stake in the outcome. Kuanta scores every company on the same framework for its sector and stage, 645+ criteria in all, in the same structure, whoever pitched and whoever is reading. It has not met the founder, so it cannot be charmed, and it has no memory of your last deal.
It also has limits, which are worth being honest about. It cannot look a founder in the eye or sense how two co-founders get on, and those things matter. An independent thesis at Rotterdam School of Management found that Kuanta and human analysts, on 150 startups, were comparably accurate and read partly different signals, and that the strongest predictor of success was the two of them agreeing.
A second opinion with no favourites
Our suggestion is to keep pattern matching and to check it. When your instinct and a structured read agree, you can move quickly and with more confidence. When they disagree, you have found the deal that deserves another hour of your time, and quite possibly the question you should have asked in the meeting.
If there is a company you felt very sure about, in either direction, send it to us and see whether the structured read agrees with you.
Questions people ask
What is pattern matching in venture capital? Judging a new startup by its resemblance to earlier successes or failures. It is a real skill, and it also overlaps with well-known cognitive biases such as recency and affinity.
How can VCs reduce bias in investment decisions? By using a consistent structure for every company, recording assessments before group discussion, and adding an independent second read that has no stake in the outcome.
Gompers, Gornall, Kaplan and Strebulaev, How Do Venture Capitalists Make Decisions? · L. Meijer (2026), Man versus machine, master’s thesis, Rotterdam School of Management
Kuanta Intelligence Team
Published on Aug 18, 2026 · Updated September 2026 · Part of the Kuanta Research & Venture Decision Science series.