Strategy & Business Modeling Series, Part 2 of 5
In Part 1 we made the case that a strategy is a hypothesis, and that modeling is what turns the hypothesis into something you can evaluate. This part takes on the dimension where that discipline is most often abandoned at exactly the moment it matters most: the financial model.
Here is the pattern we see again and again. A team does thoughtful, uncertainty-aware work on the market, the customer, and the solution — and then, when it comes time to build the financial case, they produce a single column of numbers. One market-size figure. One conversion rate. One price. One cost. The spreadsheet multiplies them together and returns a five-year revenue line that looks authoritative precisely because it's specific. It is also almost certainly wrong, and worse, it hides exactly how wrong it might be.
Why the single-point forecast misleads
A single-point forecast answers a question nobody should be asking about a new business: “What is the single result this strategy will produce?” For a mature, well-understood operation, a point estimate can be reasonable — you have history to anchor it. For a new venture, you don't. Every input is uncertain, some of them wildly so, and collapsing that uncertainty into one number doesn't remove the risk. It just hides the risk from view.
The single-point forecast fails in several specific ways. It projects false confidence — a number to two decimal places reads as knowledge even when it rests on a guess. It obscures the range of outcomes that actually matter — the difference between “this clears our threshold in every plausible case” and “this clears it only if three optimistic assumptions all land” is the entire decision, and a point estimate erases it. And it invites a quiet political failure: when a leader has to defend a single number as a personal commitment, the incentive is to pick the number that gets the project funded, not the number that's true. Finally, it anchors investment decision makers on a value that will certainly change, which can create numerous stakeholder management challenges when they do.
Model the financial logic with ranges
The fix is structural, and it follows directly from the modeling discipline in Part 1. The Financial Logic dimension of the strategy — market sizing, pricing, the revenue model, the cost structure, unit economics — is built out of inputs, and the uncertain inputs are not single values. They are ranges.
Instead of “we'll convert 4% of the segment,” the model says “somewhere between 2% and 7%, most likely around 4%.” Instead of one price, a plausible band. Instead of one customer-acquisition cost, a range that reflects what you actually know and don't. Each range is an honest statement of the current state of evidence: wide where you're guessing, narrow where you have data. As Part 1 framed it, these are the Uncertainties in the hypothesis — and expressing them as ranges is what keeps them from masquerading as facts you've already nailed down.
Two things happen when you do this. The model starts telling the truth about how much you actually know. And — this is the part teams underestimate — project leaders become far more willing to engage with the numbers, because they're no longer being asked to stake their credibility on a single figure. A range reframes the financial model from a commitment to be defended into a shared object to be evaluated and improved.
Run the ranges: Monte Carlo
Once the uncertain inputs are ranges, you can't just multiply them the way you'd multiply single numbers. You run the model many times within variance in the ranges. This is what a Monte Carlo simulation does: it samples random values from within each input's range, runs the model, records the result, and repeats — hundreds or thousands of times. Each run is one plausible version of the future, drawn in proportion to what you believe about each input. The output isn't a number; it's a probability distribution — the full shape of where the outcomes might land.
That distribution is far more decision-useful than any point estimate. It shows the spread: best cases, worst cases, and where the bulk of outcomes cluster. It shows the probability of clearing a threshold — not “the model says we'll hit the target” but “the target is cleared in roughly 70% of plausible runs.” It tells you whether the opportunity is robust (good across most of the range) or fragile (good only if several uncertain inputs all break favorably). A fragile opportunity isn't necessarily a bad one — but you want to know it's fragile before you fund it, not after.
You don't need to be a statistician or a spreadsheet wizard to do this. The mechanics are well understood and increasingly built into the tools; what matters is the shift in posture — from asserting one future to mapping the range of them.
From a distribution to a decision
A Monte Carlo distribution is genuinely informative, but on its own it's still just a picture of uncertainty. The natural next question is the one that turns the picture into action: of all these uncertain inputs, which ones actually drive the result — and which of those do we have the least evidence for? That is where the next dollar of validation work should go, and it's the subject of Part 3.
For now, the move is simpler and it's foundational: stop forecasting new businesses with single numbers. Build the financial logic out of ranges that reflect what you honestly know, run them, and look at the shape. The point of the financial model in new-business exploration isn't to predict the answer. It's to show you, clearly, how much is still in play.
BRI Associates helps companies grow by drawing on decades of practitioner experience in corporate innovation and new business development — practitioners, not pundits or academics — through direct consulting, training workshops, and Growth Forge® Software, built for the unique requirements of corporate innovation and growth organizations.
