The 23 questions every VC asks during due diligence (with the right answers)

Diligence is where rounds die.

A founder who walked out of the first call thinking it went well will get a polite pass two weeks later, and never know that diligence calls with their customers, their team, or their cap table were where the real decision happened.

This piece walks through the 23 questions a seed or Series A VC asks during diligence, in the order they typically come up. For each: what the partner is actually trying to learn, what a good answer looks like, and what a bad answer signals.

The 23 questions sort into 5 buckets: the founder, the team, the customer, the business, and the round.

The founder (5 questions)

These come up in the first or second meeting, usually woven into casual conversation. The partner is forming a view of you as a person.

1. Why this problem? Why now?

What they are learning: do you have a real reason to spend 7 years on this. Or did you pick it because it sounded good in a deck.

Good answer: a specific moment or experience that put you on this path. Not "I have always been interested in X" but "I watched my mother spend 14 hours a week on Y, and that is when I saw the gap."

Bad signal: rationalized post-hoc story that does not hold up under follow-up questions.

2. Why are you the right team for this?

What they are learning: have you shipped something hard before, and have you shipped it together.

Good answer: one specific shipped artifact per founder, plus one specific story about working together under stress.

Bad signal: heavy reliance on titles and credentials without a shipped artifact.

3. What is your relationship with your co-founder?

What they are learning: can this partnership survive 5 hard years. Most founder splits happen at year 2-3 and they kill startups.

Good answer: an honest framing of the dynamic, including how you handle disagreement. "We disagreed on X last quarter; here is how we worked through it" is much stronger than "we have always been on the same page."

Bad signal: idealized framing that suggests you have not yet had a real fight.

4. What is the worst thing about your business right now?

What they are learning: are you self-aware. Do you know your weaknesses or are you in denial.

Good answer: a specific real weakness with a specific plan for addressing it. "Our enterprise sales motion is broken; we have hired a sales lead starting in 6 weeks and the plan to fix it is X."

Bad signal: "we work too hard" or "we have too much demand to keep up." This is the most common bad answer to this question and partners notice.

5. If this does not work, what do you do?

What they are learning: how committed are you. Founders with strong fallbacks (re-join Google, take a senior role somewhere) sometimes lack the urgency that comes from no fallback.

Good answer: an honest framing of your commitment. "I am 100% in on this for the next 5 years; if it does not work I will probably try another startup" is fine.

Bad signal: equivocation or visible plan B.

The team (4 questions)

These typically come up in the second or third call, often through asks for follow-up references.

6. Tell me about your most recent hire. Why them?

What they are learning: how good is your hiring. Most early-stage failures trace back to bad early hires.

Good answer: specifics on the hiring process, the bar you set, and the trade-offs you considered. Specifics on why this person beat the alternatives.

Bad signal: "we needed someone fast." Speed-over-quality hiring at the early stage compounds badly.

7. Tell me about your worst hire.

What they are learning: same as above but the inverse. Have you fired someone you needed to. Did you do it in time.

Good answer: a specific story including how long it took you to recognize the problem and how you handled it. Honest about your own slowness if applicable.

Bad signal: "we have not had any bad hires" at year 2+. Statistically false.

8. What is the team gap right now?

What they are learning: are you self-aware, and do you have a hiring plan.

Good answer: a specific gap (sales lead, head of product, senior engineer in domain X) with a specific timeline.

Bad signal: "we have everything we need" or "we will figure it out." Both signal you have not thought about it.

9. Who are your 3 references on the founders?

What they are learning: who knows you well, and what will they say.

Good answer: 3 names, not all current investors. At least one former boss or peer who knew you before this startup. Diverse perspectives.

Bad signal: 3 references all from the current cap table. Partners will discount these.

The customer (5 questions)

These usually come during week 2-3 of diligence. Partners will call your customers directly.

10. Who are your 3 best customers and why?

What they are learning: do you know what makes a great customer for you.

Good answer: 3 names, with a specific reason each is great (use case fit, retention pattern, expansion behavior, advocacy). The pattern across them tells you who to sell to next.

Bad signal: "all our customers are great" or 3 customers with nothing in common.

11. Who are your 3 worst customers, or customers who churned?

What they are learning: do you know who you are not for. Do you fight churn or accept it.

Good answer: 3 names with the specific reason each was a bad fit, plus what you have changed about your sales process to filter them out.

Bad signal: "we have not had any churn" if your business is more than 12 months old.

12. Can I talk to 3 of your customers?

What they are learning: how strong is the customer signal. Diligence calls with customers are the single highest-weighted input in most diligence processes.

Good answer: a list of 5-6 customers, with a note on which ones to call for which signal (advocate, neutral, recently expanded). Make the intros within 24 hours.

Bad signal: hesitation. Or "let me check with them first" delays. Or curating only your most positive customers.

13. What is your net revenue retention?

What they are learning: are existing customers expanding. NRR over 110% at the early stage signals product-market fit; under 90% signals churn problems.

Good answer: the specific number with the cohort it is calculated against, plus a one-sentence narrative on what is driving it.

Bad signal: not knowing the number, or quoting gross retention to dodge the question.

14. What does the typical customer look like 12 months in?

What they are learning: is the product getting more useful with time, and is usage compounding.

Good answer: a specific cohort behavior: feature adoption, usage growth, expansion into new teams.

Bad signal: "they are happy" without specifics.

The business (5 questions)

These are the unit economics and trajectory questions. Usually asked in week 1-2 of diligence.

15. What are your unit economics today?

What they are learning: does the math work, and how confident can they be in scaling.

Good answer: CAC, LTV, gross margin, payback period. Real current numbers, not projections.

Bad signal: blended numbers, projected numbers, or numbers that include free users.

16. What is your CAC and what are you doing to lower it?

What they are learning: do you understand your acquisition mechanics, and is there a clear lever.

Good answer: CAC by channel, with a specific channel that is improving and a specific lever you are pulling on.

Bad signal: a single blended CAC with no channel breakdown.

17. What is the path to $10M ARR (or $25M, depending on stage)?

What they are learning: do you have a real model for how the business grows. Not a projection, a model.

Good answer: a specific bridge from today's number to the target, broken down by which channels and which expansion motions get you there. With sensitivities.

Bad signal: "we will scale our team and grow." This is the most common bad answer.

18. What is the biggest risk to the business right now?

What they are learning: same as the worst-thing-about-your-business question, but more specific.

Good answer: a specific risk (e.g., regulatory, competitive, technical, hiring) with a specific mitigation plan.

Bad signal: "we do not see significant risks" or generic answers like "execution risk."

19. What kills you in 18 months?

What they are learning: are you thinking about second-order risks. Most founders cannot answer this well.

Good answer: 2-3 specific scenarios that would meaningfully threaten the business, with the early indicators that would tell you each one is happening.

Bad signal: inability to think of a real failure mode. Partners read this as overconfidence.

The round (4 questions)

These come in the final stretch of diligence, typically after the partner has decided to recommend the deal to their partnership.

20. Who else is in the round?

What they are learning: who is leading, who is committing, and is there momentum.

Good answer: specifics. Who has committed, who is in active diligence, who has soft-circled. Honest about where each one stands.

Bad signal: vague references to "several other funds" without names. Partners will ask other partners about overlap; if the picture you painted does not match, trust drops.

21. What is your timeline to close?

What they are learning: is there urgency, and is the urgency real.

Good answer: a specific date with a real reason (board meeting, runway, next milestone). Not artificial pressure.

Bad signal: aggressive timeline with no real reason behind it.

22. What does this capital let you prove?

What they are learning: do you know what 18 months buys you.

Good answer: 3 specific milestones the round unlocks, with the metric for each.

Bad signal: "scale the business" or "expand the team."

23. Why us?

What they are learning: have you done the work to know why this fund specifically, or are you raising from anyone with a checkbook.

Good answer: a specific reason that is partner-specific or fund-specific. "Your portfolio company X is a likely customer for us, and we have already had two conversations with their team about a paid pilot."

Bad signal: generic reasons (great brand, helpful network) that could apply to any fund.

How to prepare

Going into diligence cold is the most expensive thing you can do. The 23 questions above are remarkably consistent across funds. Spend 4-5 hours preparing your answers before your first partner meeting. Update them after every call where a new question surfaced.

The single highest-leverage piece of prep: have a customer reference list ready before anyone asks. The first partner who asks for customer calls and gets them inside 24 hours becomes a believer. The partner who has to wait 5 days starts to wonder.

If you want a tool that runs you through a mock partner Q and A using your specific deck, Claude does this well with the right prompt (see prompt 4 in the Claude prompts piece). claudefundraiser.com builds this into the deck-scoring flow.

About Claude Fundraiser

Claude Fundraiser scores pitch decks on the 8 dimensions VCs actually evaluate (problem, market, team, traction, etc.), matches founders against 8,665 enriched investors ranked by fit, and drafts personalized cold emails grounded in the founder's own deck and each investor's thesis, in the founder's voice. Free to score and see top 10 matches. Plans at $297/$697/$1,497 for 50/200/500 drafts, or à la carte at $49/$149/$299.