strategyplaybook.aiSqwyz
Sign in
← All 124 tools
Risk · Plan

Sensitivity analysis

Varies the inputs to a financial model to see which ones actually move the answer. Usually two do and the rest are decoration.

The printable canvas and the handoff into Claude Code are part of a paid plan. See what a plan includes. It needs the Claude desktop app on this machine, and your team and the prompt library already installed in that project — we cannot see your disk, so open it there.

Stage
04 Plan
Works at
Corporate, Business
Maturity
Scaling → Large organisation
Time
Half a day once the model exists

What it is

Systematic variation of a financial model's inputs to establish which ones actually determine the outcome. Most models have twenty inputs and two that matter, and knowing which two changes both the decision and what you monitor afterwards.

Input ranges
A plausible high and low for each variable, grounded in history or evidence rather than in optimism and pessimism.
One-at-a-time analysis
Moving each input across its range with others held at base, to rank variables by their influence on the result.
Break-even points
The value at which each critical input stops the case working. More useful than any percentage sensitivity, because it is a number you can watch for.
Combined scenarios
Several important variables moving together, since real downside cases are correlated rather than independent.
Monitoring linkage
The critical inputs carried into the indicator set, so the model's assumptions are actually tracked after approval.
The mistake people makeThis tests a model; it does not build one. And a model whose inputs are all guesses will produce a confident-looking sensitivity analysis of nothing — check the inputs have some grounding before ranking them.
What it’s forThe financial case rests on assumptions and you need to know which actually move the answer.
What it’s not forYou have no model. This tests a model; it does not build one.

How you run it

  1. List the inputs and give each an honest rangeNot best case and worst case — plausible high and low, based on history or evidence.
  2. Vary one at a time firstHold everything else at base and move each input across its range. This ranks the variables by influence, which is the main output.
  3. Find the break-even on the critical onesAt what value of this input does the case stop working? That number is far more useful than a percentage swing.
  4. Then vary the important ones togetherReal downside cases involve several things going wrong at once, and they correlate more than models assume.
  5. Carry the critical inputs into monitoringThe two variables that drive the answer should appear in your leading indicators. That connection is the point of doing this.

The prompt

Two ways to run it

Run this tool in your own Claude

The short prompt starts your partner against the library on your disk. The long one carries everything with it and needs nothing installed.

🔒 Prompt locked

Everything above is free — what this tool is, what it is for, what it is not for, and how to run it as a workshop. Three tool prompts a month are free with an account; beyond three, and for the printable canvas, it is the paid part.

See plans — from £19 a month

Your playbook

It lands in the earliest stage this tool suits. Move it on the Playbook page.

You’ll need

  • A working financial model
  • A plausible range for each key input
  • Someone who will accept an unwelcome result

You’ll end up with

  • The variables ranked by influence on the outcome
  • Break-even points on the critical ones
  • The two inputs that need monitoring

Runs well with

AI operating model optionsAI risk classification and governanceAlliance and partnership designAssumption log