Scenario planning

Build three or four internally consistent futures and test each option against all of them, rather than forecasting one future.

Time cost: 1–2 days
Output: Three or four named worlds, an options grid, and early indicators.
Steps: 5

Use when

The uncertainty is structural — which world you end up in, not which number comes up.
The planning horizon is long enough that the drivers themselves could change.
A group is arguing about a forecast when they should be arguing about which world they are in.

Do not use when

The horizon is short and the drivers are stable. A range on a number is enough.
You will end up picking one scenario as the plan, which is forecasting with extra steps.

Inputs required

The decision and its horizon
The forces that could move it
Which of those are genuinely uncertain

Procedure

01
Separate certain from uncertain
Some forces are effectively predetermined over the horizon — demographics, installed infrastructure. Put those in every scenario. Scenarios differ only on what is genuinely uncertain.
02
Pick two critical uncertainties
The two that matter most and are least predictable. Two axes give four quadrants, which is the right number: three or four scenarios, no more.
03
Write each world
A short narrative per quadrant, internally consistent, with a name. It must be a world someone could live in, not a list of variables.
04
Test every option in every world
The grid is options by scenarios. Look for an option that is acceptable in all four rather than excellent in one.
05
Name the early indicators
For each scenario, what would you observe first if it were arriving? These are the tripwires that make the exercise operational rather than literary.

Characteristic failure mode

Writing one plausible scenario and two strawmen. The plan then optimises for the favoured world, and the exercise has laundered a forecast as a robustness test.

Worked example

A household deciding whether to take a large fixed-rate loan over ten years.

01Two uncertainties: income stability, and the direction of rates.
02Four worlds named and written, each a page.
03The fixed loan is fine in three and painful in one — the world with an income interruption and falling rates.
04Early indicator for that world: sector hiring freezes before rate moves.

Result

The decision changes from fixed-or-variable to fixed-with-a-six-month-reserve. That option was acceptable in all four worlds and was not on the original list.

Where to go next

Also cited by

Second-order thinking