Using AI to anticipate the 40 questions investors will ask in DD

The founders who close rounds fastest aren't the ones with the best answers. They're the ones who heard the hard questions three weeks before the investor asked them. AI is now the fastest way to surface those questions cold — and the cheapest stress test before the real room.

Why prepare for DD questions at all — won't they vary?

They vary in wording, not in category. Investors are pattern-matching against deals they've already seen. The 40 questions are roughly the same 40 questions every time; the surface form changes.

Definition: Due diligence (DD) — the structured investigation an investor runs after term-sheet interest but before final commitment, covering operating, financial, legal, and market validation of the founder's claims.

The founder who has thought through the 40 in advance walks into the call calm. The one who hasn't sounds like they're discovering their own business live, and that's the tell investors use to back away.

What are the 6 categories — with sample questions?

1. Financial (8-10 questions)

• What's your monthly burn split, gross vs net?
• Show the cohort retention curve. Where does it flatten?
• What's your CAC payback and how has it moved over 6 quarters?
• What's the assumption set behind the forecast you sent us?
• Walk us through your last re-forecast — what changed and why?

2. Customer (6-8 questions)

• Walk us through your top 5 customers — concentration, contract length, expansion history.
• What's the cancellation reason mix from churned accounts in the last 12 months?
• How many customers have multi-year contracts vs annual?
• Show us 3 customer-success calls or references we can call directly.
• What's the gross retention number — not net, gross — and the calc?

3. Team (5-7 questions)

• Walk us through the org chart and recent departures.
• Who's the most replaceable person on the team and who's the least?
• What's your hiring plan for the next 18 months and the cost?
• What's your compensation philosophy — base, equity, bonus mix?
• Are any key people on visas, and what's the renewal timeline?

4. Product (6-8 questions)

• Walk us through the roadmap and what's been killed in the last 12 months.
• What's the engineering velocity trend over the last 4 quarters?
• Show us the analytics — DAU, feature adoption, time to value.
• What's the build vs buy split — what's homegrown vs vendor?
• Where's the technical debt and what's the cost to clear it?

5. Market (5-7 questions)

• Walk us through your competitive map and the wins and losses.
• Who are the 3 competitors that worry you most and why?
• How is the market changing in the next 24 months and what does that mean for the moat?
• What's the realistic TAM serviceable in the next 5 years (not the consultant TAM)?
• Where are you positioning vs the incumbent — disruption, displacement, or wedge?

6. Legal (4-6 questions)

• Walk us through the cap table including unvested grants.
• Any open litigation, claims, or threatened claims?
• Walk us through IP ownership — is everything assigned correctly?
• Any change-of-control clauses in customer or vendor contracts?
• What's your data privacy and AI Act compliance status?

That's roughly 40 questions across 6 categories. The exact count varies by stage — earlier rounds skip some financial detail; later rounds add more legal depth.

Copy/paste question-bank prompt

You are a sceptical Series B investor reviewing an SMB for due diligence.
You will receive a one-page company brief. Your job: generate the 40
substantive questions you would ask in DD, organized by category.

Categories (with target question count):
- Financial (8-10)
- Customer (6-8)
- Team (5-7)
- Product (6-8)
- Market (5-7)
- Legal (4-6)

Rules:
- Each question must be specific to this company, not generic.
  E.g. NOT "what's your churn?" but "your investor update mentioned
  churn rose to 3.2% in Q2 — walk us through what drove it."
- Each question must be answerable in 2-5 minutes by a prepared founder.
- Include at least 5 "trap" questions: ones where the obvious answer is
  wrong and the right answer requires the founder to disclose a complication.
- Group questions logically inside each category (don't randomize).
- Do NOT generate answers. Founder will answer cold.

Output: numbered list per category, ready for founder mock-pitch session.

The trap-question rule is what separates a useful question bank from a generic one. Real DD has trap questions; a question bank without them is a softball list.

Tool tip (AIAdvisoryBoard.me): The reason most founders are unprepared for DD isn't that the questions are hard — it's that the daily operating reality doesn't surface them. A founder running Plan → Fact → Gap visibility across the company sees the gaps months in advance — the churn cohort that's drifting, the engineering velocity that's slowed, the customer concentration that's growing. By DD time, the trap questions aren't traps anymore. The 7-day diagnostic shows where today's gaps are: https://aiadvisoryboard.me/?lang=en.

How do you run the practice-pitch workflow?

Generate the 40 questions. Then run three sessions.

Session 1: Cold answers (founder solo, 90 minutes)

• Founder answers each question into a voice recorder
• No prep, no looking up numbers
• Tag answers by confidence: green (clean), yellow (rough), red (no answer)
• Red and yellow are your prep list

Session 2: Prep the reds and yellows (founder + finance/legal as needed, 3-4 hours)

• Each red answer: dig into the data, draft a real answer, write it down
• Each yellow: rehearse until green
• This is where the data room work pays off — every answer should trace to a doc

Session 3: Mock pitch with AI as sceptical investor (founder, 45 minutes)

• Feed AI the full 40-question bank
• Ask AI to play the investor — pick 12 questions at random, ask follow-ups based on answers
• Founder answers verbally, AI probes weaknesses
• Re-prep anything still rough

By session 3, the real DD call becomes a recital. That's the goal.

Good vs bad question generation

Bad (generic): "What's your customer acquisition cost?"

Good (sourced from brief): "Your Q1 update said CAC payback was 14 months. The Q2 update is silent on payback. What's the current number and what shifted?"

Bad (no trap): "What's your team like?"

Good (with trap): "You hired 12 people in Q1. Your Q2 plan was for 8 more. The team page on your site shows 4 net adds since Q1. Walk us through what happened to the rest."

The good versions show the investor has actually read the materials. The founder has to be ready for that level of specificity.

Manager scan (2-minute digest example)

• Plan: DD prep complete 2 weeks before first investor call. Target: 40 questions answered, 3 mock pitches done.
• Fact: Session 1 cold answers — 22 green, 11 yellow, 7 red.
• Gap: 7 red answers cluster in financial (3) and team (2) — finance reforecasting and recent founder-VP friction need prep work.
• Plan: Session 2 prep with finance review on financial reds.
• Fact: 2 of 3 financial reds resolved with finance data; 1 required acknowledging a real weakness (churn calc methodology shift Q3 to Q4).
• Gap: Decide whether to volunteer the methodology shift or wait for it to be asked. Volunteering builds trust; waiting risks looking like cover-up.
• Plan: Session 3 mock with AI as investor.
• Fact: AI's follow-ups exposed 2 new yellows that weren't on the original list — both around competitive positioning.
• Gap: Add competitive deep-dive prep before real calls.

Micro-case (what changes after 7-14 days)

A 50-person vertical SaaS founder had a strong narrative and weak DD muscle. Three previous Series A attempts had stalled in DD; investors went quiet after the second call. The founder ran the 6-category workflow with AI two weeks before re-engaging investors. Session 1 surfaced 9 red answers — most around customer cohort behavior and a complicated bridge round in the cap table. Two weeks of prep, three mock pitches, and the actual investor DD calls felt rehearsed in a good way. The lead investor commented to the founder that this was "the most prepared DD pitch I've sat through in a quarter." The round closed.

Note on this case: This example is illustrative — based on typical patterns we observe with companies of 30-500 employees, not a single named client. Specific numbers are rounded approximations of common ranges, not guarantees.

Tool tip (AIAdvisoryBoard.me): Founders who close rounds fast usually have one thing in common: the gaps DD will probe are already visible in their own dashboards weeks before the round starts. Plan → Fact → Gap surfaces operating drift daily — churn cohort drift, hiring vs plan delta, customer concentration trend — so the founder hears the hard questions from their own data before any investor asks. See https://aiadvisoryboard.me/?lang=en.

FAQ

Won't different investors ask different questions? The wording varies; the underlying categories are stable. Series A investors lean financial and team; Series B leans customer cohorts and unit economics; growth-stage leans market and competitive. The 40-question structure scales — just shift the category weights.

Should I rehearse the answers word-for-word? No — that's how founders sound canned. Rehearse the structure of each answer (lead with the number, give the context, name the risk) and let the words come naturally. Investors can tell scripts.

What if AI generates a question I literally cannot answer? That's the most valuable output. Either you can answer it after prep, or you can't — and either way, finding out two weeks before the call instead of during it is the win.

Can AI play the investor for the full mock pitch? For the first round of mocks, yes — it's free, fast, and patient. The second round of mocks should be a live human (an advisor, a friendly investor) who reads body language and tone. AI catches content gaps; humans catch delivery gaps.

Conclusion

DD doesn't have to be the part of fundraising where rounds die. Six categories, roughly 40 questions, three sessions of prep — and the real call becomes a rehearsal. AI can generate the questions, play the sceptical investor, and surface the gaps weeks before they cost real money.

Brief AI on your company in one page. Generate the 40. Answer cold. Prep the reds.

If you want a system that surfaces Plan → Fact → Gap continuously — so the questions investors ask are questions you've already been answering for yourself — see how the 7-day diagnostic works at https://aiadvisoryboard.me/?lang=en.