Experienced angels screen deals for a handful of recurring patterns: founders with deep knowledge of their market, fast iteration, early evidence that customers pull the product, and a clear reason the timing works now. Pattern-matching misleads when it rewards resemblance to past winners, learns only from survivors, or trusts confident instincts that years-long feedback never tested.
Pattern recognition is the skill investors talk about most and examine least. Every angel who has seen enough decks starts to feel a deal is good or bad within minutes. Sometimes that feeling is real expertise compressed into intuition. Sometimes it is a bias that has never been checked. The useful question is not whether to use patterns, but which ones deserve trust.
What patterns do experienced angels screen for?
The patterns that come up again and again in how seasoned angels describe their screening fall into a few groups.
About the founders
Founder-market fit. Founders who know the problem from the inside, through work in the industry, years as the customer or a technical insight others lack. The guide to evaluating founder-market fit breaks this down.
Speed of learning. How much the founders have changed their thinking since the last conversation, and how quickly they ship and test.
Clarity under questioning. Founders who can say what they don't know, and what would change their mind, rather than defending every point.
References from people who have worked with them. Former colleagues and early customers are hard to fake.
About the business
Pull rather than push. Users who come back, customers who pay before the product is finished, a waitlist that grows without spending.
Why now. A recent change in technology, cost, regulation or behavior that makes the idea possible today when it wasn't before.
A market that can get large. Not just a big number in a slide, but a credible path from the first customers to many more.
About the round
A plan to the next milestone. Enough money to reach the evidence the next investor will need.
Clean structure. A sensible cap table, reasonable terms and no unexplained history.
None of these is new, and that is the point. The value is not in the list but in how accurately an investor can judge each item in a short meeting. A decision framework for angel investing turns lists like this into a repeatable process.
Which patterns does the evidence support?
Some patterns have published data behind them.
Investing in industries you know. In Robert Wiltbank and Warren Boeker's 2007 study of angels in groups (Kauffman Foundation), investments connected to the angel's own industry expertise returned noticeably higher multiples.
Founders with a prior success. Paul Gompers, Anna Kovner, Josh Lerner and David Scharfstein (Journal of Financial Economics, 2010) found that venture-backed founders whose previous company went public were more likely to succeed in their next venture than first-time founders or founders whose previous company failed.
Doing the work. In the Wiltbank and Boeker data, exits where angels spent more than the median time on diligence returned far more than those where they spent less.
The numbers are in what the data says about luck and skill.
Where does pattern-matching mislead?
Feedback comes too late to train intuition. Daniel Kahneman and Gary Klein's paper "Conditions for Intuitive Expertise" (American Psychologist, 2009) argues that whether to trust an intuitive judgment depends on how predictable the environment is and whether the person has had the chance to learn its regularities. They also note that feeling confident is not a reliable sign of being right. Early-stage investing is a hard case: outcomes take years to arrive and are heavily affected by luck, so an angel's gut gets little clean feedback.
Resemblance gets mistaken for quality. A founder who looks, talks or studied like the founders of past winners feels familiar, and familiar feels safe. PitchBook's All In reports have put all-female founding teams at about 2% of US venture dollars (2022). Whatever the full set of causes, numbers like that are a reminder that "looks like past winners" is a filter that can exclude good founders and the returns they would bring.
Survivors teach the wrong lessons. If you study only the companies that succeeded, every trait they share looks like a cause. Many failed companies had the same traits. Patterns learned from winners alone overstate how predictive those traits are.
The last cycle sets the template. Patterns formed in one market, such as cheap capital or a particular platform shift, can be wrong in the next. Angels who learned in a boom often over-weight growth and under-weight capital efficiency.
Early success is partly place and time. Ramana Nanda, Sampsa Samila and Olav Sorenson found that venture firms' early IPO successes predicted later ones mainly because those wins brought better deal access, not because the firms had a lasting edge in picking. An angel who did well early may be reading skill into good timing.
How do you build pattern recognition that holds up?
Write predictions down. For every deal you invest in and every one you pass on, note what you expect and why. Without a written record, memory rewrites the reasons to fit the outcome.
Track your passes. The companies you declined that later did well are the most useful data you have about your own patterns.
Separate familiar from good. When a deal feels right quickly, ask which part feels familiar and whether that part actually predicts success.
Look for the disconfirming fact. Before investing, name the one thing that would make you pass, and check it.
Lean on domain knowledge. Patterns are most reliable in markets you understand from the inside, which is also where the published data shows better outcomes.
Volume is what supplies examples. My portfolio has 234 angel investments, and Lenny's Newsletter ranked me the most active angel investor of 2024 and tenth over five years. Volume like that gives an investor many examples, but it does not by itself show which instincts are right; only a written record compared against outcomes does. More on that is in the lessons from making hundreds of angel investments and the piece on spotting startups with 100x potential. The angel investing section covers evaluation and diligence in more depth.
Frequently Asked Questions
Q:How many deals does it take to develop pattern recognition?
There is no fixed number, and seeing many deals is not the same as learning from them. What matters is getting feedback: tracking what you predicted against what happened, including for companies you passed on. Because outcomes take years, expect pattern recognition to develop over a long period.
Q:Is gut feeling reliable in angel investing?
Gut feeling is most reliable in sectors where you have deep experience, and least reliable where you don't. Research on intuitive expertise shows that confidence alone is not evidence of accuracy, especially when feedback is slow and noisy. Use intuition to decide what to investigate, then check it with diligence.
Q:What is the difference between pattern recognition and bias?
Pattern recognition links a trait to outcomes because the trait actually predicts them. Bias links a trait to outcomes because it resembles past winners or feels familiar. The way to tell them apart is to test your patterns against your own record, including the deals you passed on.
Q:Should new angels rely on other investors' patterns?
Learning from experienced investors is useful, especially by co-investing and asking why they decided as they did. But borrowed patterns carry their owner's biases and their market's conditions. Start in a sector you know and build your own record alongside.
Q:Do patterns differ by sector?
Yes. Signals that matter in consumer products, such as retention and organic growth, differ from those in enterprise software or deep technology, where customer contracts or technical milestones carry more weight. Patterns are most reliable inside the sector where they were learned.