A business plan is a list of assumptions dressed up as a spreadsheet

Business planning · Assumptions

Over the years, I have read and built a fair number of business plans. Ambitious ones, mostly; cautious ones, rarely. They all had one thing in common: people talked a great deal about their numbers, and in the end very little about what held those numbers up. The aim of this article is to offer you a change of perspective: stop reading a business plan as a numerical prediction, and start reading it as what it really is — a list of assumptions dressed up as a spreadsheet. Or a set of bets on the future.

The numbers are only consequences

Say you are preparing the launch of a drug. Your business plan proudly announces: 50,000 diagnosed patients, 20% of them treated with your product, 12 packs per patient per year. That makes 120,000 packs sold every year. Your spreadsheet will hand you that result with two decimal places if it has to. It may be wrong, but it will have two decimal places.

Where do those three numbers come from? The 50,000 assumes the patient population exists, is diagnosed, and is reachable. The 20% assumes prescribers will choose your product — and that reimbursement will follow. The 12 assumes patients will take their treatment all year long, without interruption. Three bets. The 120,000, on the other hand, is not information: it is the arithmetic consequence of three bets.

And this is true of every cell in the file. Year-3 EBITDA is probably one of the most scrutinized and least informative numbers there is: a stack of dozens of bets, multiplied by one another. As long as those bets stay implicit, the plan looks solid. Which is precisely what should worry you.

The cell test

The right reflex fits in one question. Take any number in the plan and ask: what has to be true for this cell to tell the truth? You will see things surface that the spreadsheet does not show:

• 20% annual growth assumes your competitors will watch you do it without reacting;
• a stable margin assumes your suppliers will not raise their prices — ask them what they think;
• a spring launch assumes everything that needs to be ready beforehand will be, which is already a dozen cascading assumptions.

You will notice, in passing, that the most dangerous assumptions are usually not in the spreadsheet at all: the approval that never arrives, the key person who leaves, the single supplier that shuts down. The spreadsheet only computes what it is given.

An assumption worthy of the name is falsifiable

Once the assumptions are extracted, you still have to write them properly. Most assumptions found in business plans are unassailable — and that is exactly the problem. "The market is buoyant" cannot be contradicted by reality; the sentence risks nothing, and therefore teaches you nothing. Compare:

• "The market is buoyant" → "at least 50,000 patients are diagnosed each year, and one in five is eligible for our treatment";
• "Prescribers will follow" → "20% of the target specialists have prescribed the product at least once by the end of month six";
• "We will be on the market quickly" → "the price is published and reimbursement effective within twelve months".

The left-hand version is a slogan. The right-hand version can be contradicted by reality — which is exactly what we ask of it. A well-formulated assumption contains a number, a deadline, and a failure condition: if any of the three is missing, you are holding an opinion, not an assumption.

A business plan does not die from its calculations. It dies from its assumptions.

Not all assumptions are equal

Your list will quickly grow to fifteen or twenty lines, and you will have neither the time nor the money to verify everything. Two questions are enough to sort it: if this assumption breaks, does the plan survive? And: how much does checking it cost? A critical, uncertain assumption gets tested right away. A critical but well-documented assumption gets monitored. A secondary assumption goes on the list and waits its turn.

And to test it, there is no need to commission a five-figure market study. As with data analysis, the minimum data necessary is enough. Want to check the adoption bet? France's public health-insurance databases will tell you how many patients are already treated, and with what. It is free, it is a little dry, and it is worth all the slides in the world.

Track your assumptions the way you track a budget

Surfacing and testing is good. But a business plan lives on, and its assumptions age — some well, some badly. The discipline that changes everything is modest: at every monthly or quarterly review, reread the list. For each assumption: still true, weakened, or contradicted. Green, amber, red. Ten minutes.

The day a critical assumption turns red, the decision — continue, pivot, stop — is no longer a matter of ego or internal politics: it has been written into the list from the start. That discipline — surface, then test, then track — is what separates a plan you endure from a plan you steer. It is, as it happens, the heart of our approach.

Next time someone hands you a business plan, resist the urge to ask "how much?". Ask instead: "what has to be true?".