By stage
Pre-seed diligence
How to assess a company with no revenue, no product and nothing to measure — where the diligence is almost entirely about the founders and the problem.
Realistic time Four to eight hours over two weeks
At pre-seed there is nothing to analyse. There may be a prototype, there is rarely revenue, and any financial model is a set of assumptions arranged in a spreadsheet. Angels who try to run seed-stage diligence at this stage produce elaborate answers to unanswerable questions and mistake the effort for insight.
What can be assessed is real, though: who the founders are, whether they understand the problem better than you do, whether they can attract other people, and whether the thing they are building could plausibly become large. Those four questions are the whole of pre-seed diligence, and they take longer to answer honestly than a spreadsheet does.
The other thing to accept is the base rate. Most pre-seed companies fail, and no amount of diligence changes that. Diligence at this stage is not about avoiding failure — it is about avoiding the specific failures that were visible in advance: a founding team that will not survive its first disagreement, a market that cannot support a venture outcome, a problem the founders have not actually experienced.
01
The founders
You are underwriting people for eight to twelve years. That is longer than most marriages last and considerably longer than most jobs, and the thing you are trying to establish is whether these particular people will still be doing this when it stops being exciting.
The most useful signal is depth of engagement with the problem. Founders who have lived the problem — as an operator, a customer, a practitioner — talk about it differently from founders who identified it in a market map. They know the unglamorous details, they have opinions about things you did not think were contentious, and they can tell you why the obvious solution does not work.
Founder relationships deserve direct attention. Ask how they met, how long they have worked together, what their worst disagreement has been and how it resolved. Co-founder breakdown is among the most common causes of early-stage failure, and a pair who have never disagreed have usually not been tested rather than being unusually compatible.
Check
- How did the founders come to this problem? Lived experience or market analysis?
- How long have they worked together, and in what capacity?
- What is the equity split, and is everyone on a vesting schedule?
- Has anyone already left the founding team, and what did they keep?
- Can they explain what they got wrong in the last six months?
- Who have they already persuaded to join, advise or use the product?
02
The problem and the market
The test is not whether the market is large but whether it could plausibly support a company worth a hundred times what you are paying. That is a different and more useful question, and it can usually be answered with arithmetic rather than a research report.
Be suspicious of top-down market sizing. A slide claiming a $50bn market almost always describes a category rather than an addressable opportunity. Bottom-up sizing — how many customers exist, what would each pay, what share is winnable — is harder to produce and far more informative, and founders who have done it are demonstrating something about how they think.
The strongest signal at this stage is usually why now. Something must have changed to make this possible or necessary — a regulation, a technology, a cost curve, a behaviour. Founders who cannot answer that question convincingly are often building something that has been tried before and failed for reasons that still apply.
Check
- Bottom-up market sizing: how many customers, paying what?
- Why is this possible now and not three years ago?
- Who has tried this before, and why did they fail?
- What does the customer do today instead? Doing nothing counts.
- How would the company reach its first hundred customers?
03
Evidence, however thin
There is usually more evidence at pre-seed than founders think to show you. Waiting lists, letters of intent, unpaid pilots, a prototype that people use without being asked, a community, a newsletter with real engagement — none is revenue, and all of it is signal.
What matters is whether the evidence was hard to get. A hundred sign-ups from a launch post is a distribution result, not a demand result. Five customers who agreed to pay before the product existed is a demand result, and it is worth more than a hundred times the sign-ups.
Ask to see the product, and ask to use it if that is possible. Founders who are reluctant to show you a rough product are usually reluctant for a reason, and roughness at pre-seed is expected and forgivable in a way that evasiveness is not.
Check
- Use the product yourself, however early it is.
- Ask what evidence was hardest to get, and why.
- Talk to one or two users or prospective customers if the founders will introduce you.
- Distinguish between people who signed up and people who came back.
04
The structural basics
Even at pre-seed there are a handful of structural things that are cheap to check now and expensive to discover later. All of them are fixable at this stage and close to unfixable at Series A.
The recurring ones: intellectual property that has not been assigned by contractors or by an employer the founder has just left; a departed co-founder holding a large stake; equity promised verbally and never documented; and, for US companies, 83(b) elections that were never filed within the 30-day window.
Check
- Is all IP assigned to the company, including from contractors?
- Does any founder have restrictive covenants with a previous employer?
- Is founder vesting in place and documented?
- For US companies: did every founder file an 83(b) election within 30 days?
- What convertible instruments are already outstanding, and at what caps?
Stop and think
Red flags
- A founding team that has never disagreed, or cannot describe a disagreement honestly.
- A co-founder who has already left holding a substantial vested stake.
- No vesting on founder shares, and reluctance to put it in place.
- Market sizing that is entirely top-down, with no bottom-up arithmetic behind it.
- An inability to say what would have to be true for the company to fail.
- Evasiveness about outstanding convertibles or promised equity.
- A founder who has not spoken to a potential customer in the last month.
The other side of the table
What founders are taught about this
Kevin Hale - How to Evaluate Startup Ideas
The framework YC teaches founders for judging whether an idea is worth pursuing — which is the same judgement you are making from the other side of the table, with the same information.
Published by Y Combinator. Channel verified from ycombinator.com, and YouTube's own oEmbed response names that channel as this video's author. Pressing play loads content from YouTube.
Take these into the room
Questions to ask the founders
- What did you believe six months ago that you no longer believe?
- What would have to be true for this to fail?
- Which part of the plan are you least confident about?
- What have you already tried that did not work?
- How much have you and your co-founders each put in, in cash?
- What are you going to spend this round on, specifically?
- What does the next round need to be true for you to raise it?
From the decoder