The five questions deal teams skip.

Why the important questions get skipped

Commercial diligence runs on a clock, and the data room is optimized for the clock. Everything in it is an answer to a question the seller expected. That is not deceit; it is preparation. The gap is structural: the buyer asks what the deck makes easy to ask, and the deck was assembled by people who know what the deck will be asked.

The five questions below share a property. Each one can only be answered by going underneath a summary number to the transactions, the records, or the people that produced it. That takes a day or two each, which is why they get skipped, and it is also why they are worth asking. The full checklist has 47 of these across ten dimensions; these are the five we would run first if we were given one week.

1. What is gross retention when computed from invoices, not CRM fields?

Every deck has a retention number. Almost none of them were computed from billing. They come from a CRM field, a spreadsheet someone maintains, or a dashboard whose logic nobody in the room can explain. CRM fields drift: a downgraded account stays "active", a churned logo is still "customer" because nobody flipped the status, a renewal that shrank by half is recorded as a renewal.

What to ask for: the invoice ledger for the last eight quarters, by account. Rebuild gross retention cohort by cohort from what was actually billed. Then compare it to the slide. The gap between the two numbers is the first real fact of the diligence, and how the management team reacts to the gap is the second.

Red flag: the two numbers cannot be reconciled and nobody is surprised.

2. How much of the pipeline is aged past the median sales cycle, or simply dead?

Pipeline coverage is quoted as a ratio: three times quota, four times quota. The ratio counts everything with a dollar value and a stage. It does not tell you how old those opportunities are, and age is the whole story. An opportunity that has sat open for twice the median cycle is not pipeline. It is a record nobody wanted to mark lost, because marking it lost changes a number someone is measured on.

What to ask for: the open-opportunity export with created dates and last-activity dates, plus the median cycle length from the last two years of closed-won deals. Age every open deal against that median. Then recompute coverage counting only committed stages and only deals younger than the median. That is the coverage the thesis is really standing on.

Red flag: a large share of pipeline has no activity logged in the last quarter, and the forecast still includes it.

3. Where does founder gravity still close the deals, and what happens without it?

This is the question people are most reluctant to ask out loud, because the founder is in the room and the answer is often "everywhere". Founder-led selling is not a flaw at the right stage; it is the engine. The diligence question is whether anything else has been built alongside it, because the thesis usually assumes the founder steps back, and the revenue model rarely models what steps back with them.

What to ask for: the last twenty closed-won deals with the full contact history. Count how many had the founder in the deciding conversation. Then ask the reps, one at a time and without the founder present, to walk through their last win. If the walkthrough keeps routing through one person, you have your answer.

Red flag: the sales team can describe the process, but cannot name a significant deal that closed without founder involvement.

4. What would organic visibility look like if paid spend stopped tomorrow?

The marketing slide shows channels and pipeline sourced. It rarely separates what is bought from what is earned. Paid spend can produce a healthy-looking top of funnel right up to the day it stops, and a buyer who inherits a pipeline that only exists while the invoices to the ad platforms keep clearing has bought a subscription, not an asset.

What to ask for: attribution you can rebuild yourself, not dashboard totals: the raw lead-source data joined to closed deals. Separate paid, organic, referral, partner, and outbound. Then look at the organic column on its own. Search visibility for the terms buyers actually use, whether AI assistants can describe the company accurately, referral flow from customers. If that column is thin, the marketing engine is a spend line, and the model should treat it as one.

Red flag: the team cannot say which channels produce pipeline that closes, only which channels produce leads.

5. Can the CRM revenue picture be reconciled to billing, and when did anyone last try?

The quiet one. Every number in the commercial deck comes from the CRM. Every number in the financial deck comes from billing. Diligence teams read both and almost never put them side by side, because each team owns one and the tie-out belongs to nobody. When the two disagree, and they usually do, the size of the disagreement is a direct measure of how much the company knows about its own business.

What to ask for: closed-won bookings by month from the CRM, and invoiced revenue by month from billing, for the same period. Reconcile them. Ask when the company last did this itself. The answer "we do it every month" with evidence is a green flag worth more than most of the deck. The answer "we have never needed to" tells you the reporting the business runs on has never been tested against reality.

Red flag: lead-source and stage fields are overwritten by every tool that touches the CRM, so even the CRM does not agree with itself.

Running all five in a week

• Day one: request the invoice ledger, the open-pipeline export with dates, the closed-won history with contacts, and the raw lead-source data. Ask for exports, not dashboards.
• Days two and three: rebuild retention and coverage from the raw data. Two numbers, two gaps against the deck.
• Day four: the rep conversations, without the founder. Founder gravity and organic reality both surface here.
• Day five: the CRM-to-billing reconciliation, and a single page: five questions, five findings, what each one does to the model.

The point

None of these questions are exotic. They are skipped because each one costs a day and produces a number that might complicate a thesis everyone is already fond of. That is precisely the argument for asking them before the wire rather than in the first board meeting after it. The other 42 questions matter too. These five are where the surprises live.