STUPIDPRO.Opening your desk…

Scenario analysis for founder and research decisions

Explore consequences under explicit assumptions without presenting a scenario as a forecast.

For Public,Founders,Researchers · Reviewed 2026-10-09

A scenario asks what would happen under specified conditions; a forecast makes a claim about what is likely to happen. Keep those roles distinct. Central-bank stress testing, such as the Bank of England’s programme, illustrates the use of hypothetical conditions to examine resilience. A startup’s planning worksheet is a much smaller exercise with its own assumptions and limitations.

Our suggested starting point is one decision: hiring, entering a market, launching a feature or renewing a supplier contract. Define the planning horizon and the variables that actually affect the outcome. For a software business, these might include retention, acquisition cost, usage cost and payment timing. Give every assumed input a label and keep observed inputs in separate columns.

Build a base case and contrasting cases that change a few meaningful drivers. Do not assign probability weights merely to make the output look sophisticated. Calculate the implications consistently, then test whether the proposed action remains workable across cases. A sensitivity table can show which single assumption changes the conclusion most, helping the team decide what to research next.

Finally, connect each scenario to observable signposts and a review date. Decide in advance what evidence would prompt a change in spending or scope. Preserve the assumptions when updating the worksheet so the team can distinguish new evidence from a changed model. The valuable result is a decision with explicit conditions and a learning plan, rather than a confident-looking spreadsheet that hides uncertainty behind one revenue number.

Questions to investigate

  1. Which assumption most changes the decision when varied?
  2. What observable signpost would cause the team to revise the plan?
  3. Are these hypothetical cases being mistakenly presented as probable outcomes?

Primary sources

Related research

All research guidesResearch workspace