Framework library · Decision making and problem solving
Decision tree analysis
A decision tree turns a choice under uncertainty into branches: the options you control, then the outcomes you do not, each with a probability and a value. Rolling the values back gives each option an expected value, and the tree shows which uncertainty would change the answer.
Use it when
- You face a choice whose outcome depends on events you cannot control, such as demand, a regulator or a competitor.
- Options differ in timing, for example act now or wait for information, and the value of waiting needs to be shown.
- You need to show a board why the option with the biggest upside is not necessarily the best.
Avoid it when
- You cannot put even rough probabilities on the outcomes. Use scenario planning instead.
- The downside of an option is unacceptable whatever its expected value. Expected value assumes you can live with any outcome.
- There are dozens of interacting uncertainties. A Monte Carlo model handles those better.
How to run it
Write the options as the first branches
Include doing nothing and waiting. Options are things you choose.
Add what could happen after each
Two to four outcomes per option, each with a chance. The chances under one option must add up to 100%.
Value each end point
Net present value is best; a consistent profit or cost measure will do. Include the cost of the option itself.
Roll back
The expected value of a set of outcomes is their chance-weighted average. The best option is the one with the highest expected value. The tree below does the arithmetic.
Test the probabilities
Change the chance you are least sure of until the best option changes. If a small change flips it, the decision rests on that number.
Work through it
Answer the questions below, or load the worked example to see a finished one. The drawing updates as you type. Export the result as a PowerPoint deck, a Word document, an Excel workbook, a PDF or plain text.
What you type stays in this browser, so you can close the page and come back to it. It is not sent to Blue Prysm or anyone else, and the exports are made here, on your device. Privacy policy.
Mistakes to avoid
- False precision: probabilities to the nearest per cent that nobody believes.
- Leaving out the option to wait, which is often worth most when uncertainty will resolve soon.
- Choosing by expected value when one branch would bankrupt the business.
Where it comes from
John F. Magee, "Decision Trees for Decision Making", Harvard Business Review, July to August 1964; formalised as decision analysis in Howard Raiffa, Decision Analysis: Introductory Lectures on Choices under Uncertainty (1968). Source.
Use it with
Further reading
Work through it with us
The frameworks here are free to use as they stand. If you would rather work through the question behind this one with us, these are the ways an engagement starts.