What-if scenario

A what-if scenario is a copy of your financial model with one or more assumptions changed (a hiring freeze, a price increase, a lost customer), used to see how results would differ from the base case. It answers a decision question, not a forecasting question: not “what will happen” but “what would this do.”

How to run a what-if properly

Three things have to be fixed before the answer means anything. State the base case and the version it came from. Change one thing, or a named bundle of things that belong together, and leave everything else alone. And decide in advance which output lines you are watching, so the scenario is judged on cash runway or operating margin rather than on whichever number looks best afterward.

Second-order effects are where most what-ifs quietly fail. Freeze hiring and payroll falls, but so does the capacity that fed the revenue plan, and the recruiting spend, and eventually the support cost that scales with customers. If those links are not in the model, the scenario reports a saving with none of the cost attached.

Why what-ifs go stale

A scenario built as a duplicated spreadsheet starts drifting the moment the base case updates. Two weeks later, actuals have landed, the base plan has been re-versioned, and the scenario is being compared against a base case that no longer exists. The fix is structural: keep scenarios as a scenario branch tied to the same underlying model rather than as a separate file.

The related trap is scenario sprawl. Ten variants with names like “v3 final revised” and no record of what changed in each is worse than three well-labeled ones. Each scenario should carry its changed assumptions as an explicit list.

What good looks like

A useful what-if states its base case and version, lists exactly which assumptions moved, shows the delta on the outputs that matter, and can be re-run against fresh actuals without being rebuilt. When it turns into a decision, the assumptions become something to track against, not something to forget.

The value of a scenario depends entirely on whether dependent figures recompute correctly. Rexfin lets you describe a scenario in plain language and watch every dependent line update deterministically against the same governed definitions and cited actuals as the base case, so the comparison holds.