The seven sections every business plan gets interrogated on

A business plan is not judged on what it says. It is judged on what a sceptical reader can do to it. Investors, lenders, boards and acquirers all run some version of the same interrogation, whether they write it down or not. Our engine formalises that interrogation into seven canonical sections. If your plan holds up in all seven, it will survive most rooms. If it fails in one, that one failure tends to set the price of everything else.

1. Executive summary

The test is falsifiability. Can the reader extract, from the first page, a claim that could in principle be proven wrong? A summary that says who pays, for what, at what price, and why they keep paying is falsifiable. A summary built from mission language is not, and experienced readers treat unfalsifiable openers as a signal about everything underneath. The most common failure is the vision-first summary: three paragraphs about the future of an industry, no number a diligence analyst could check. The second most common is the vanity aggregate, a cumulative lifetime figure placed where a current-period figure should be. Both get flagged in minutes.

2. Market and customer

The test is construction. Any market-size number gets rebuilt from its inputs, and most cannot be. A total addressable market produced by multiplying a big population by an assumed per-head spend is not evidence; it is arithmetic on hope. The interrogation asks narrower questions. Who is the actual buying unit? What do they spend today, on what substitute? What share of that spend is realistically contestable in the plan's own timeframe? Claims of having no competitors receive special attention, because they are almost always a category error: the competitor is whatever the customer does now, including nothing. A plan that names its substitutes and prices against them reads as written by someone who has met the market.

3. Product and moat

The test is replication cost. For every claimed advantage, the reader asks what it would cost a funded competitor to copy it, and how long the copy would take. Most stated moats fail this test because they are descriptions of activity, not barriers: we have a platform, we have data, we have a brand. Stacking five weak moat claims does not make one strong one; it usually signals that the author knows none of them holds alone. A defensible moat section quantifies one mechanism, such as switching costs measured in observed retention, or a cost position traceable to something structural, and shows the evidence. One quantified moat beats five asserted ones.

4. Go-to-market

The test is whether growth is an input or an output. Plans in trouble treat the growth rate as a decision: we will open this many locations, hire this many reps, acquire this many users per quarter. Plans that survive treat growth as the output of a unit that already works, with a stated gate: expansion proceeds when a cohort hits defined economics, and halts when it does not. The interrogation looks for the gate. It also looks for the arithmetic connecting acquisition cost to customer value, cohort by cohort, with the definitions written down. Undefined terms are treated as concessions. If the plan says a mature location earns a given margin and never defines mature, the reader assumes the definition moves whenever the number needs help.

5. Financials and unit economics

The test is reconciliation. Every derived or adjusted figure must reconcile to the statutory numbers, line by line, and any metric invented by the company gets read as an admission of what the standard metrics show. Custom profitability measures that exclude major operating costs are the classic case; the reader simply re-adds the exclusions and prices the result. The second test is duration: how the timing of money in compares with the timing of money committed. Long fixed obligations funded by short cancellable revenue is a structural exposure, and calling that structure an advantage does not change what it does in a downturn. The third test is the downside case. A financials section with no stress scenario is treated as a financials section with a hidden one.

6. Team and operations

The test is constraint. Not whether the team is impressive, but whether anything constrains it. The reader looks for concentration: one person carrying execution, technical judgement and the investor relationship at once; voting structures that make the founder unremovable; related-party arrangements with no independent sign-off. Track-record claims get the denominator question: previous successes mean little without the full list of previous attempts. The operations half of the test asks whether the plan's stated pace is physically achievable by the described organisation, at the described headcount, with the described suppliers. A growth cadence that outruns the operating apparatus is not ambition; it is an undisclosed hiring plan.

7. Risk and compliance

The test is quantification. Listing risks is not managing them, and a risk register that enumerates six existential threats without attaching a probability, an exposure or a mitigation to any of them is decoration. The interrogation cross-references this section against the rest of the plan: risks disclosed here and reframed as advantages elsewhere, insurance and governance gaps that contradict the team section's confidence, regulatory exposure concentrated in exactly the markets the revenue depends on. The sharpest question is the simplest: which single risk, if it lands, ends the company, and what in the plan shows it has been priced rather than merely mentioned?

Why seven, and why in this order

The sections are ordered the way a reader's trust erodes. A weak executive summary lowers the prior for the market section; an unconstructed market number lowers the prior for the moat; and by the financials, the reader is no longer evaluating claims, they are auditing them. Plans do not usually die from one bad section. They die from a compounding discount applied across all seven. The purpose of an adversarial review before the real audience sees the plan is to take that discount in private, where it is cheap, instead of in the room, where it is priced.

Our engine runs five independent red-team roles across these seven sections and synthesises their verdicts into a KEEP / MODIFY / KILL matrix, three ranked fatal-flaw candidates with the test that would confirm each, and three next actions. You can read three full anonymised reports on real public plans from the landing page, or see how the flaws play out in practice in Three fatal flaws that killed real business plans.