Methods · 01/10

Scenario Analysis

A structured method for developing several plausible pictures of the future in order to prepare and test strategic decisions under uncertainty.

Purpose

Scenario analysis does not produce a forecast; it produces a set of internally consistent pictures of possible futures. It therefore answers a different question than classical planning — not what is most likely to happen, but within what space the future can plausibly unfold. Originating in the military planning work of Herman Kahn in the 1950s and brought into corporate strategy by Pierre Wack at Shell, it has become the core instrument of strategic foresight.

Its real value lies less in the finished scenarios than in the process itself. Assumptions about the future are made explicit, critical uncertainties are separated from mere trends, and strategies can be tested against several conceivable environments rather than a single preferred one. Organizations gain a shared language for uncertainty — and a basis for defining early indicators that reveal which direction the world is actually taking.

Procedure in five steps

Frame the focal question
The process starts with a precise focal question: which decision is being prepared, for which business, over what time horizon? A scope that is too broad yields arbitrary scenarios; one that is too narrow misses relevant environmental forces. Horizons of five to fifteen years are common, depending on the investment cycles of the industry.
Collect driving forces
Environmental factors are then gathered systematically — typically along the PESTEL dimensions, complemented by industry- and market-specific forces. Sources include studies, expert interviews, patent data and weak signals from environmental scanning. The result is an ordered list of factors with an initial assessment of their impact.
Identify critical uncertainties
Each factor is rated by impact and uncertainty. High-impact, low-uncertainty factors enter all scenarios as trends; high-impact, high-uncertainty factors become the critical uncertainties that differentiate the scenarios. This separation is the analytical core of the method.
Construct the scenarios
Consistent combinations of the uncertainties' possible outcomes are assembled — classically through a consistency analysis, or in simplified form through a two-axes matrix. Three to four scenarios have proven practical: clearly distinguishable, free of internal contradiction, and written up as narratives with a name, a logic and a development path.
Derive implications and indicators
Finally, the organization's strategy is wind-tunnelled through each scenario: which options hold up everywhere, which only in one future? The output is a set of robust moves, contingency plans and early indicators whose occurrence signals which scenario is gaining probability.

Strengths and limits

Expands the space of thinking beyond an extrapolated present and makes discontinuities discussable.
Forces implicit assumptions into the open and separates reliable trends from genuine uncertainties.
Allows strategies to be stress-tested against several environments, distinguishing robust options from risky ones.
Creates a shared language for uncertainty that keeps working in management long after the analysis itself.

The process is demanding — a careful cycle ties up expertise from several functions over weeks.
Scenarios carry no probabilities; reading them as forecasts misunderstands the method.
Group processes tend to converge toward the middle, softening extreme but relevant futures.
Without continuous upkeep, scenarios age quickly and decay into a one-off workshop exercise with no effect.

With AI and agents

The most labor-intensive part of scenario work — collecting and assessing driving forces — changes fundamentally with AI. Instead of a one-off research phase, agents can observe the environment permanently: they scan studies, news flows, patent filings and draft regulation, map findings to the defined driving forces, and report when the assessment of a factor shifts. The annual scanning project becomes a continuous process with a far broader base of sources.

Scenario construction gains depth as well. Language models check bundles of assumptions for internal contradictions, propose overlooked combinations of uncertainty outcomes, and draft first scenario narratives that the team then sharpens. The connection to early indicators is particularly powerful: agents monitor the defined signals continuously and trigger an update as soon as a scenario gains or loses plausibility. Scenarios thus move from static documents to maintained, living working states.

Human judgment remains irreplaceable at three points: choosing the focal question, deciding which uncertainties truly matter for the organization, and drawing the strategic conclusions. Language models also tend to extrapolate the plausible past — the surprising, structure-breaking futures still require imagination and the willingness to take uncomfortable possibilities seriously.

Relation to scenarios

Scenario analysis is the center toward which the other tools converge: PESTEL supplies the driving forces, Five Forces the industry logic, SWOT the internal view. Once the method is understood, the genuinely interesting step follows — developing one's own pictures of the future for one's own business: with a concrete focal question, the critical uncertainties of the specific environment, and scenarios against which real decisions can be tested.

Generate your own pictures of the future at picturesoftomorrow.com

How ready is your organization for AI and agents? — Take the AI Readiness Check