VC Due Diligence Checklist: What Investors Ask For, Why They Ask, and What Weak Answers Signal (Pre-Seed to Series B)

Q: What is VC due diligence?

VC due diligence is the verification process an investor runs between expressing interest and closing: confirming that what a company claimed is true, identifying what was not disclosed, and deciding whether the remaining uncertainty is acceptable at the price. It spans legal and corporate documents, commercial evidence, and judgment about the team and market.

Q: What is included in a VC due diligence checklist?

Two layers. The legal and corporate layer covers minutes, charter documents, capital stock, IP assignment, employment agreements, and debt instruments. This is what Cooley GO's Sample VC Due Diligence Request List and Y Combinator's Series A diligence checklist cover. The commercial layer covers team, product and technology, market, traction, business model, and risk. Most published checklists cover only the first layer.

Q: What are the biggest red flags in startup due diligence?

Unassigned intellectual property, a metric whose definition changes between the deck and the data room, the claim of 'no direct competitors', growth that stops when paid spend stops, and an inability to name the principal risk specifically. Unwillingness to share evidence that plainly exists is more concerning than the absence of evidence that does not exist yet.

Author: Eric Levine, Founder of StratEngine AI | Former Meta Strategist | UCLA Anderson MBA

Published: July 28, 2026

Reading time: 8 minutes

Summary

Most published VC due diligence checklists are document request lists. Cooley GO's Sample VC Due Diligence Request List is organized into eight legal categories — Actions and Minutes, Charter Documents, Capital Stock, Legal and Regulatory, Intellectual Property, Management and Employees and Consultants, Debt Financing, and Other Agreements — with two columns beside each item: "Date Provided" and "Does Not Exist." Y Combinator's Series A diligence checklist, attributed to Jason Kwon, General Counsel of YC Continuity, serves the same purpose: get the data room assembled before the term sheet so the closing process is not held up hunting for documents.

Both are excellent at what they do. Neither tells you what the answers mean.

This checklist covers the other half — the commercial diligence that sits alongside the legal file, why each item gets asked, and what a weak answer signals to the person reading it. It is staged by round, because diligence depth changes sharply between a pre-seed and a Series B.

This is not legal advice. The legal and corporate items belong with counsel. Use Cooley's or YC's list for those, and a lawyer to interpret them.

How does VC due diligence change by stage?

The most common founder mistake is preparing for the wrong depth. The same question means different things at different rounds.

Pre-seed

There is almost nothing to verify — no revenue history, often no product in market, no cohort data. Diligence at this stage is mostly founder diligence: who are these people, have they done hard things before, and do they understand the problem more deeply than an outsider would after a week of reading. What gets checked: founder backgrounds and claims about them, cap table cleanliness, IP assignment (did the technical founder build this while employed somewhere else), and whether the founding team is complete enough to build the thing.

Seed

Early evidence exists but is thin and mostly founder-reported. Diligence starts triangulating: do the usage numbers match what the analytics actually show, do customer references corroborate the narrative, is the growth organic or bought. What gets added: product usage data, early retention, pipeline quality, and a real market-sizing conversation rather than a slide.

Series A

This is where diligence becomes genuinely rigorous. There is enough data to test claims properly, and enough capital at risk to justify the effort. What gets added: cohort retention curves, unit economics with the assumptions exposed, customer concentration, churn analysis, sales efficiency, a full legal and corporate file, and reference calls with customers and former colleagues.

Series B

The questions shift from "does this work" to "does this scale, and what breaks when it does." What gets added: multi-year cohort behavior, margin structure at scale, competitive displacement evidence, org and hiring plan credibility, and a much harder look at whether the market is as large as the earlier rounds assumed.

Team diligence checklist

Founder backgrounds and verifiable credentials. Asked because the team is the only thing at pre-seed with any track record attached. Weak answer: credentials that cannot be independently confirmed, or a résumé framed to imply seniority the dates do not support. Vagueness about what someone actually did in a prior role is more concerning than an unimpressive prior role.
Founder commitment and equity split. Asked because uneven commitment tends to surface later as founder conflict, which is expensive to resolve after money is in. Weak answer: a co-founder who is part-time with no defined date for going full-time, or a split that does not reflect actual contribution and has never been discussed openly.
Why this team for this problem. Asked because domain insight is hard to fake and hard to acquire quickly. Weak answer: a market chosen because it is large rather than because the founders understand something specific about it.
Key-person dependency. Asked because concentration risk at seed becomes structural risk later. Weak answer: one person holds all the technical or customer knowledge and there is no plan to change that.

Product and technology diligence checklist

What exists today versus what is roadmap. Asked because decks routinely present planned functionality in the present tense. Weak answer: a demo that cannot be run without the founder driving it, or features described without a clear statement of what is shipped.
Technical architecture and key dependencies. Asked to understand what the company actually controls. Weak answer: core functionality that is a thin wrapper on a third-party service the company has no relationship with or leverage over.
IP assignment and ownership. Asked because unassigned IP is one of the few genuinely deal-breaking findings. Weak answer: contractors without signed assignment agreements, or code written while a founder was employed elsewhere without a release. Cooley's list covers the documents; the judgment call is whether the gaps are curable before close.
Security and data handling. Asked with weight proportional to what the product touches. Weak answer: handling sensitive customer data with no articulated practice around it.

Market diligence checklist

How the market size was derived. Asked because top-down sizing is the most common analytical failure in a deck. Weak answer: a large number multiplied by an assumed percentage, with no bottom-up build from customer count times realistic contract value.
Why now. Asked because timing explains why the opportunity is available rather than already taken. Weak answer: a "why now" that would have been equally true five years ago.
Competitive landscape, including the honest version. Asked because how a founder describes competitors reveals how carefully they have looked. Weak answer: "no direct competitors." It almost always means the founder has not looked hard, has defined the category narrowly enough to be trivially true, or is not counting the status quo — and the status quo is usually the real competitor.
Where the company loses. Asked because every company loses some deals. Weak answer: an inability to name a segment or use case where a competitor is genuinely the better choice.

Traction diligence checklist

Metrics with their definitions and measurement windows. Asked because the definition frequently does more work than the number. Weak answer: "users" without saying whether that is registrations, activated accounts, or people who came back. A metric whose definition shifts between the deck and the data room is a serious signal.
Retention and cohort behavior. Asked because retention is among the hardest metrics to manufacture. Weak answer: aggregate growth presented without cohorts, which can hide flat or declining retention behind new acquisition.
Organic versus paid. Asked to distinguish demand from purchased attention. Weak answer: growth that stops when spend stops, presented without that context.
Customer concentration. Asked because concentration is both revenue risk and a signal about repeatability. Weak answer: a large revenue share from one or two logos, especially where those relationships are personal to a founder.
Third-party corroboration. Asked because founder-reported numbers are inputs, not findings. Weak answer: no analytics access, no payment processor data, no customer references willing to talk. Note that "unverified" is not the same as "false" — early companies legitimately have little third-party evidence. The concern is unwillingness to provide what does exist.

Business model diligence checklist

Unit economics with assumptions exposed. Asked because the assumptions matter more than the outputs. Weak answer: a CAC or payback figure with no visible derivation, or one that excludes categories of cost that are plainly part of acquisition.
Whether margins improve with scale. Asked because venture returns depend on economics that get better, not merely hold. Weak answer: a cost structure that scales linearly with revenue, described as though it does not.
Pricing rationale. Asked because pricing reveals how well the company understands the value it delivers. Weak answer: pricing set by looking at competitors, with no evidence of testing what customers will actually pay.

Risk diligence checklist

The principal risk, named specifically. Asked because a founder who cannot name the thing most likely to kill the company either has not thought about it or is managing the conversation. Weak answer: generic risks — competition, execution, hiring — offered in place of the specific one.
Regulatory exposure and its timeline. Asked because regulatory risk tends to arrive later and larger than founders expect. Weak answer: a regulated-adjacent business with no articulated view on how the rules might move.
What would have to be true for this to be a zero. Asked because it is the fastest way to find out whether a founder has genuinely stress-tested their own thesis. Weak answer: an answer that avoids the question.

How long does VC due diligence take?

It varies widely by round and by fund, and any single number would be misleading. Pre-seed decisions can move quickly because there is little to verify. Series A diligence takes longer, because there is finally enough data to test properly and enough capital at risk to justify the effort. The practical lever founders control is preparation: an assembled data room removes the delay that comes from hunting for documents after a term sheet is signed. That is the argument YC's checklist makes explicitly, and it is why Cooley publishes its request list as a document founders can work through in advance.

Where does AI help in due diligence, and where does it not?

AI is genuinely useful for the mechanical layer: extracting every claim from a deck, checking public sources for corroboration, flagging where a stated number conflicts with another stated number, and producing a consistent structure across every deal so that the tenth memo of the week is as rigorous as the first.

What it does not do is exercise judgment. Deciding that customer concentration is acceptable because the concentration is with a customer who is themselves scaling — that is a human call. The failure mode worth naming is treating a general-purpose summary as though it were verification. A summary reorganizes what the deck said; it does not check whether any of it is true. That distinction is the difference between summarizing text and analyzing a business, and it is exactly where screening tools miss the outlier.

For a worked example of the separation done properly, this sample investment memo keeps claims from the deck and independently researched findings in separate sections, then scores each area with the reasoning left visible.

Frequently Asked Questions

What is VC due diligence?

It is the verification process an investor runs between interest and closing: confirming that what a company claimed is true, identifying what was not disclosed, and deciding whether the remaining uncertainty is acceptable at the price. It spans legal and corporate documents, commercial evidence, and judgment about the team and market.

What is included in a VC due diligence checklist?

Two layers. The legal and corporate layer — minutes, charter documents, capital stock, IP assignment, employment agreements, debt instruments — is what Cooley's and YC's published lists cover. The commercial layer covers team, product, market, traction, business model, and risk. Most published checklists cover only the first.

How is due diligence different at pre-seed versus Series A?

At pre-seed there is little to verify, so diligence concentrates on founders, cap table cleanliness, and IP ownership. By Series A there is enough data to test claims properly: cohort retention, unit economics with exposed assumptions, customer concentration, reference calls, and a complete legal file.

What are the biggest red flags in startup due diligence?

Unassigned IP, a metric whose definition changes between the deck and the data room, "no direct competitors," growth that stops when paid spend stops, and an inability to name the principal risk. Unwillingness to share evidence that plainly exists is more concerning than the absence of evidence that does not exist yet.

Sources

About the Author

Eric Levine is the founder of StratEngine AI. He previously worked at Meta in Strategy and Operations, where he led global business strategy initiatives across international markets. He holds an MBA from UCLA Anderson. He has direct experience building AI-powered strategic analysis tools used by consultants, executives, and venture capitalists to generate data-driven framework analysis and institutional-grade strategic recommendations in minutes.