Narratives, not forecasts
A scenario is a structured, internally consistent story about how the external environment could plausibly develop over the life of the decision. It carries no probability and makes no claim to be right; its job is to be coherent enough that the team can walk the decision through it. That distinction is what separates scenario planning from forecasting. A forecast concentrates attention on the single future the model considers most likely. A scenario set deliberately spreads attention across several, because for long-lived commitments the most likely future is still, individually, unlikely.
When it beats point forecasting
- External uncertainty is high and structural: regulation, technology substitution or geopolitics could move the ground, not just the numbers.
- The commitment horizon is long: infrastructure, decade-scale outsourcing, market entry, anything whose payback outlives the current planning cycle.
- The cost of being confidently wrong exceeds the cost of being roughly prepared, which is the usual shape of irreversible decisions.
The two-axes method
List the uncertainties that could change the decision, then select two that are high-impact, genuinely uncertain and largely independent of each other. Crossing them produces four quadrants; each becomes a named scenario with a short narrative. As an illustration, take a regional grocer weighing a ten-year commitment to an automated fulfilment centre, with online grocery share and labour-market tightness as the axes.
| Scenario |
Online demand |
Labour market |
What the investment looks like here |
| Full Shift |
Accelerates |
Tightens |
The facility is the business; the real risk was building only one |
| Patient Automation |
Plateaus |
Tightens |
Payback is slower but sound: automation still beats scarce labour |
| Crowded Race |
Accelerates |
Eases |
Volume arrives, but rivals running cheap labour compress the margin |
| Yesterday's Bet |
Plateaus |
Eases |
The stranded-asset case; staging and exit terms decide the damage |
Using all four, not picking one
The exercise fails at the last step more often than at any other: teams build four scenarios and then optimise the decision for the one they privately expect. The discipline is the reverse. Walk the committed decision through each quadrant and record where it breaks; redesign for robustness where the cost is low (staging, exit clauses, optionality) and accept named exposure where it is not. Then attach signposts: observable early indicators that say which scenario is arriving, each with someone whose job is to watch it.
How scenario work decays
- All four scenarios are variations of the present with the dials nudged, so nothing the team currently believes is ever put at risk.
- The set is the official future plus three strawmen built to lose, and the exercise ends up ratifying the plan it was meant to test.
- Scenarios are written, presented and filed: no signposts, no owner, no trigger for revisiting the decision when the world picks a direction.
- The axes chosen are the two things the team finds most interesting rather than the two uncertainties that would actually change the decision.
What operators who lived the discontinuity add
Planning teams treat discontinuities as theoretical because, for most of them, they are: careers are long, but rarely long enough to include a supply shock, a regulatory reversal and a platform collapse. Selected senior operators have usually lived through at least one of the futures the team is only writing about, and they remember which early signals were visible, which were dismissed, and what the organisations that moved early did differently. Digital Advisory exists to put that lived range into the room before commitment: a confidential brief to the Global Board is, in practice, a way of having your scenarios read by people who have been inside them.
