The model that wins a term sheet is not the same model that survives diligence. We see this every quarter. A founder gets to a signed term sheet on a deck and a workbook that looked fine in a pitch, and then the buy-side analyst opens the file and the conversation changes shape entirely.
What we see
The models that fail diligence in the first sitting share a small number of recurring problems.
The revenue build is bottom-up but the assumptions are top-down. The headline number reconciles to a sensible market share, but the line-by-line drivers — pipeline, conversion, average contract value, ramp — do not produce the same number when you actually multiply them out. The reviewer notices in twenty minutes.
The cost base is anchored on the most recent run-rate and grown by a flat percentage. There is no step function for the second sales hire, no FX assumption for the London office, no detail on how the cost of goods responds to volume. When the diligence team asks what happens at half the revenue, the model has no answer.
The cash bridge does not close. Operating cash, working capital, capex and financing each live on their own tab, and the linkages have been done by hand. Someone has overwritten a formula with a hard number to make a quarter balance. The reviewer finds it on the second pass.
What works
The models that hold up under serious diligence tend to do three things, and they do them deliberately.
They separate inputs from logic from output. There is one tab where assumptions live, one tab where the calculation happens, and one tab that prints the numbers. No mixing. A reviewer can change a single assumption and watch the entire model respond, which is what they will want to do.
They build the revenue line in the same units the business actually operates in. If the business sells annual contracts, the model should think in annual contracts. If it sells projects, it should think in projects. A SaaS-style ARR build for a project business is a flag.
And they document the working capital and cash logic in writing, on the model itself. A short note next to the cash flow statement that explains how DSO is calculated, what is in working capital and what is not, and where the financing assumption sits, is worth more than thirty pages of supporting paper.
A model that does these three things rarely wins a higher valuation on its own. But it is the difference between a deal that closes on the agreed terms and a deal that gets repriced in the last fortnight.
