The Seed Diligence Checklist

Seed diligence has a sizing problem in both directions. Too little, and you're funding a pitch. Too much, and you're running growth-equity process on a company with nine months of history — losing the deal to faster firms while learning nothing the extra weeks could actually teach you. This checklist is sized for reality: the material questions answerable in one to two weeks, in rough priority order.

Team — where most of the answer lives

At seed, you are underwriting people more than numbers.

• Founder-market fit. Why are these specific people unusually suited to this specific problem? Lived experience with the pain beats credentials pointed vaguely at the space.
• References — the real ones. Two or three from the founders' list, then two or three back-channel from your own network. The scripted references calibrate the story; the unscripted ones move decisions. Ask every reference the same closer: would you work for them again, and would you invest?
• The founding team's history together. How long, in what configuration, through what adversity? Co-founder rupture remains one of the most common seed-stage killers.
• Equity split and vesting. A wildly lopsided split or absent vesting schedule is a live grenade — surfaced now, it's fixable at close.

Market — honest sizing, not slide sizing

• Is the problem urgent for a definable buyer, or merely real?
• Bottom-up TAM only. Number of plausible customers times plausible price. Top-down billions from an analyst report are decoration.
• Why now — what changed in technology, regulation, or behavior that makes this fundable today when it wasn't three years ago?
• Who dies if this works: which incumbents' revenue does the company eat, and how hard will they fight?

Product and traction — signal, not scale

Seed traction is about slope and quality, not magnitude.

• Use the product yourself. Non-negotiable.
• Talk to three to five customers or users directly. What you're listening for is pull: would they be genuinely upset if the product disappeared? Polite enthusiasm is a no.
• Retention and engagement in whatever form exists — cohort curves for consumer, usage depth and expansion for B2B, waitlist-to-active conversion for pre-launch.
• How much was learned per dollar burned to date. Capital efficiency at seed predicts capital efficiency later.

Cap table and round — the mechanical killers

• The full SAFE and note stack, with every cap and discount. Stacked SAFEs convert all at once at the priced round, and founders are routinely surprised by their own dilution. Model the conversion; know what the founders will actually own at Series A.
• Prior investors: who's in, at what terms, and is anyone notable conspicuously not following on?
• Dead equity — departed founders or advisors holding meaningful unvested-free stakes.
• Round construction: size, instrument, valuation, your allocation and resulting ownership, option pool timing (pre- or post-money), and pro rata rights.

Legal basics — an afternoon, not a month

• Clean Delaware C-corp (or local equivalent) in good standing.
• IP assignment from every founder and contractor — the single most common seed-stage legal defect, and the one Series A counsel will definitely find.
• No encumbrances from prior employers of the founders (non-competes, moonlighting IP claims).
• Any regulatory exposure specific to the sector, flagged even if unresolved.

The output: a one-page memo

Force the work into a decision document: thesis in two sentences, the three things that must be true, the two biggest risks and why you're accepting them, the honest bear case, and entry ownership against your construction model. If the memo won't converge in two weeks, that itself is the finding — at seed, an unresolvable diligence question usually means the answer is no, or the answer is conviction, and either way the checklist has done its job.