Due diligence for angel investors is the work you do to find out whether a startup is real before you wire money. At a minimum, you check five things: the founders, the market, the product and its traction, the deal terms and cap table, and references you source yourself. The data shows it pays off.
Angels who spent more than 20 hours on diligence saw a 5.9 times average return, against 1.1 times for those who spent less. That gap comes from the largest study of angel returns ever run, and it reframes the whole exercise. Diligence is the one lever, inside your control, that separates the angels who make money from the ones who do not.
The question that decides every angel deal: is this real?
A pitch is built to impress you. The deck is the best ten slides out of fifty. The model is a set of assumptions arranged to slope upward. The references the founder offers are the people who will say nice things. None of that is dishonest. It is just what a sales document is.
Your job as an angel is to get underneath the document and determine which claims in it would survive contact with reality. Did revenue actually land in the bank, or is it a signed letter of intent that may never convert? Is the founder the person who built the thing, or the person who can talk about it? Does the market want this, or does one friendly customer want this?
This is the proof problem, and it sits at the center of every angel deal. The founder has information you do not. Paper cannot close that gap, because paper is curated. Only diligence closes it.
The data: diligence hours map directly to angel returns
The defining study here is Returns to Angel Investors in Groups, run by Robert Wiltbank and Warren Boeker with funding from the Ewing Marion Kauffman Foundation. It remains the largest dataset of accredited-angel outcomes assembled: 3,097 investments by 538 angels, with data on 1,137 exits and closures.
The headline number is encouraging. The average return across the portfolio was 2.6 times the money in about 3.5 years, a 27 percent internal rate of return. That beats most public-market benchmarks over the same window.
The number underneath it is the one that should change your behavior. When the researchers split deals by how much diligence the angel did, the gap was enormous. Deals where the angel spent fewer than 20 hours on diligence returned 1.1 times the money on average, barely break-even. Deals where the angel spent more than 20 hours returned 5.9 times. Push past 40 hours and the multiple climbed higher still. As the Angel Capital Association summarizes the work through its education partner Seraf, more diligence time, domain expertise, and active involvement after the check all tracked with better outcomes.
There is a reason this matters so much for individuals. Nearly 70 percent of those angel investments returned less than the money put in. Angel returns are a home-run game: a few outliers carry the whole portfolio. You cannot pick the winners reliably, but the data says you can shift the odds with hours. Diligence is how you stop overpaying for the 70 percent that go nowhere.
What kills startups is your diligence checklist in disguise
If you want to know what to check, study what actually goes wrong. CB Insights has catalogued hundreds of startup post-mortems, and the same handful of causes lead the list every time. Running out of cash is the most cited. Underneath it sits the absence of a real market, what founders call product-market fit, which appears in roughly four of every ten failures. Bad timing, broken unit economics, and getting outcompeted fill out the rest.
Read that list again as a set of questions. Does anyone actually want this (product-market fit)? Is now the moment, or are they two years early (timing)? Does each sale make money, or lose it (unit economics)? Will the cash last long enough to find out (runway)? Every failure mode on that list is something you can probe before you wire, while you still have the choice. Your diligence is most valuable when it is pointed at the things that kill companies. The decorations on a deck can wait.
The five checks an angel actually runs
You are an individual investor. You do not have analysts, a hundred-hour process, or the leverage to demand a full data room before a first meeting. So angel diligence comes down to doing the highest-leverage checks well, within the time you have. Five areas carry most of the weight.
1. The founders
Professional investors are nearly unanimous on this. In the survey behind How Do Venture Capitalists Make Decisions?, Paul Gompers at Harvard Business School and his co-authors found the team was the factor VCs named most often behind both their wins (96 percent) and their losses (92 percent). Nothing else came close.
Track record is part of the picture. Data compiled by Crunchbase and Equidam puts first-time founder success around 18 percent, while founders with a prior win land closer to 30 percent. The takeaway is calibration. A prior win lowers your risk but never removes the need to check. With a first-time founder, you are underwriting raw ability and coachability, so demand more direct evidence of both.
The practical checks: read everything they have written or said publicly. Map their LinkedIn history against the story they tell you. Then call people who worked with them, including people the founder did not put on the reference list. The back-channel reference is the most useful 30 minutes in angel diligence and the one amateurs skip.
2. The market and the wedge
A great team in a market that does not exist still loses. You are checking two things: is the market real and growing, and does this company have a wedge into it. Look for evidence of demand the founder did not manufacture. Inbound interest, organic signups, customers who found them, a waitlist that grew without paid ads. Be skeptical of top-down "if we capture one percent of a huge market" math, which the failure data shows is where timing and product-market-fit deaths hide.
3. The product and real traction
If the founder claims revenue, verify every dollar. Ask for the payment processor dashboard itself, since a slide proves nothing. Distinguish signed revenue from pipeline. Talk to two or three actual customers and ask the only question that matters: what would you do if this product disappeared tomorrow. The answers separate a nice-to-have from something people depend on.
A recent NBER working paper by Xiaoyong Fu and Lucian Taylor that analyzed 21,000 deals found that 95 percent of venture deals showed no detectable in-person diligence at all, and that diligence drops 22 percent when the investor is busy with their existing portfolio. Even professionals cut this corner under time pressure. As an angel writing your own check, the customer call is the one piece of work you should never outsource.
4. The terms and the cap table
You can love a company and still get a bad deal. Sanity-check the valuation against the market: Carta put the median seed pre-money valuation at $16 million in 2025, and noted that nearly one in five venture rounds early in the year were down rounds. If the ask is wildly above market for the stage and traction, that is a flag.
Then read the cap table, the running list of who owns what slice of the company, before you fall in love. Who already owns how much? Is there a messy stack of earlier SAFEs, the agreements that convert into equity in a later round, waiting to dilute your ownership? Is the option pool, the shares set aside for future hires, sized sensibly? Are there unusual liquidation preferences from a prior investor that put you behind in a sale? For a small check, a light read is enough: know what you are buying and where you sit in the stack. For the deeper version of this, our guide to cap table management for startups and the founder-facing financial due diligence walkthrough cover the mechanics.
5. References and the back-channel
Founder-supplied references confirm the founder is likeable. The references you find yourself confirm whether the founder is honest. Talk to former co-workers, earlier investors, and ideally a customer who churned. One reframed question does most of the work: "If you were investing your own money, would you back this person again, and what would you watch out for?" The pause before the answer tells you as much as the words.
Angel diligence plays by different rules than VC diligence
Angel diligence differs from VC diligence in three ways: scale, time, and edge. An angel writes $10,000 to $100,000 solo in roughly 20 hours, while a fund deploys $1 million or more with a team over 40 to 100 hours. The biggest mistake new angels make is importing an institutional checklist they cannot actually execute. A fund and an individual are playing different games, with different constraints. Here is how the two compare.
Dimension | Institutional VC | Individual angel |
|---|---|---|
Typical check | $1M to $15M+ | $10,000 to $100,000 |
Diligence time per deal | Often 40 to 100+ hours, a team | The 20+ hours Wiltbank links to 5.9x returns, usually solo |
People doing the work | Analysts, partners, outside counsel | You |
Access to data | Full data room, management access | Whatever the founder shares, often a deck and a call |
Financial depth | Bank reconciliation, 409A, ARR build | Spot-check revenue, runway, cap table |
Real edge | Process, proprietary data, brand | Speed, domain knowledge, judgment, the back-channel |
Biggest risk | Over-processing, missing the window | Under-diligencing, relying on others' conviction |
Doing less is exactly what hurts angel returns, so the goal is to spend your scarce hours where an individual has an edge: judgment about people, a domain you know well, and references no fund will bother to chase. Pay for the things that require a specialist, like a legal read of unusual terms. Skip the theater of pretending you can run a 100-hour fund process alone.
The proxy-diligence trap, and the proof layer that fixes it
There is a comfortable shortcut that quietly destroys angel returns: letting someone else's conviction stand in for your own. A named lead is in, so you follow. A fund you respect wrote a check, so you assume the work is done. Researchers call this proxy due diligence, and it is everywhere in angel deals.
The trouble is that the people you are copying often did less work than you think. The same VC survey found the average firm screens about 200 companies to make just 4 investments a year, and that over 30 percent of deals arrive through professional networks while only 10 percent come inbound from the company. Conviction travels through networks faster than evidence does. And remember, 20 percent of those VCs do not even forecast cash flows. The lead you are following may be pattern-matching rather than diligencing. You are the one writing the check, so the conviction has to be yours.
This is the gap SeedForge was built to close, from the founder's side. A founder runs one 30-minute AI session that walks through exactly what an investor probes in the first few meetings: the team, the market, the traction behind the numbers, the use of funds. The output is a Living Profile, a structured, shareable proof document that connects real data, the deck, and the founder's own answers in one link. When a founder sends you that link before your call, you arrive already knowing what is real, so the 20 hours the data says matter go to judgment instead of fact-gathering. It removes the part of diligence that is just chasing basic facts, so your hours land where your edge actually is, and your read begins one level deeper. SeedForge does not score the company for you, and it does not replace your judgment. You still make the call, now with proof in front of you.
A practical angel diligence checklist
Use this as the floor and build up from there. Wiltbank's 20-hour threshold is the line where returns start to separate, so treat it as a budget to hit.
Founder back-channel: at least two references you sourced yourself, beyond the founder's own list.
Revenue verification: payment-processor or bank evidence for every dollar of claimed revenue.
Two customer calls: the "what if it disappeared" question, asked plainly.
Market reality: one piece of demand evidence the founder did not manufacture.
Cap table read: ownership, prior SAFEs and notes, option pool, liquidation preferences.
Valuation sanity: the ask compared against the stage benchmark.
Terms review: a specialist read on anything unusual in the documents.
Your own write-up: one page on why this clears the bar and what could kill it.
If you want to see what founders are increasingly bringing to that first call, the way diligence is moving across the ecosystem is covered in our pieces on how investor matching platforms compare and the broader seed stage due diligence standard.
Where diligence is heading
The tooling is changing fast. Industry surveys aggregated in Affinity's 2026 guide show 55 to 72 percent of VC firms now use AI somewhere in their diligence process, and one fund cut its initial screening from 45 minutes to 8 minutes per company. The same shift is reaching angels through platforms like AngelList and data tools like Crunchbase. The work that used to take hours of fact-gathering is compressing toward minutes.
Diligence stays just as important. The tooling only moves where the value sits. When the facts arrive structured, the gathering is commoditized and the judgment becomes the edge. The angel who wins in 2026 lets the machine handle the gathering and spends their human hours on the judgment the machine cannot make: is this founder the real thing, and is this market about to move.
So set a 20-hour budget, point it at the five checks above, and write the check only once the proof holds up. That is the whole job.
Frequently Asked Questions
How much due diligence should an angel investor do?
Aim for at least 20 hours per deal. The Kauffman Foundation study by Wiltbank and Boeker found angel investments with under 20 hours of diligence returned 1.1 times the money on average, while those with more than 20 hours returned 5.9 times. Diligence hours are the lever most directly inside your control.
What should an angel investor check before writing a check?
Five areas: the founders and their track record, the market and the company's wedge into it, the product and verified traction, the deal terms and cap table, and references you source yourself. Point your time at what kills startups, like weak product-market fit and broken unit economics. The polish on a deck rarely tells you whether the company survives.
How is angel due diligence different from VC due diligence?
A fund deploys $1M or more with analysts, a data room, and a multi-week process. An angel writes $10,000 to $100,000, usually solo, with limited access. The angel edge is judgment, domain knowledge, and back-channel references rather than process depth. Copy the rigor and skip the institutional machinery you cannot run alone.
What are the biggest red flags in angel diligence?
Revenue that cannot be verified in a payment processor, a valuation far above the stage benchmark, a messy cap table stacked with prior SAFEs, founder-supplied references only, and a founder who deflects direct questions. Any single one is a conversation. Several together is a pass.
Should I just follow a lead investor instead of doing my own diligence?
No. Relying on someone else's conviction is called proxy diligence, and it quietly hurts returns. The average VC firm screens 200 companies to fund 4, and conviction spreads through networks faster than evidence. You are writing the check, so the diligence has to be yours, even if a respected lead is already in.
Can a startup's diligence materials speed this up?
Yes. When a founder shares a structured proof document, such as a SeedForge Living Profile, you arrive at the first call already knowing what is real. That lets your diligence hours go to judgment instead of basic fact-gathering. It removes the slowest, lowest-value part of the work while you still make the decision.
About the Author
David Rakusan spent 7 years on the investor side before founding SeedForge. He has evaluated and helped deploy capital across more than 100 startups, holds an INSEAD MBA, and is a CFA charterholder. SeedForge is the proof layer for early-stage fundraising: founders prove their startup is real in one 30-minute session, and investors arrive at the first meeting already knowing what holds up.