Strategy & competition
Scenario Planning
A discipline for building a small set of plausible, structurally different futures and rehearsing strategy inside each one. The aim is not prediction but re-perception: loosening the official future that management quietly assumes, before events do it the hard way.
Also known as Scenario analysis, Intuitive logics scenarios, Shell scenario method. First set out by Herman Kahn; Pierre Wack and colleagues at Royal Dutch Shell; Peter Schwartz in 1985; the primary source is cited in full below.
Where this is contested
No single originator: Herman Kahn developed scenarios for military futures at RAND and the Hudson Institute in the 1950s and 1960s, Pierre Wack and Ted Newland adapted them for business at Shell from 1971, and Peter Schwartz and Global Business Network codified the 2x2 workshop method most practitioners now use.
- Format
- Mapping
- Level
- Corporate · Business unit
- Best for
- Analyse the environment · Assess risk · Evaluate options
- Decision stage
- Diagnose · Explore options · Review
- Difficulty
- Advanced
- Time to apply
- Two workshops of a day each at minimum, with research weeks between them; serious corporate exercises run three to six months.
Plate · The model
The components
Frame the focal question
The decision-anchored question and time horizon that give the exercise its boundary. Everything downstream, which forces count, which uncertainties are critical, is judged relative to this frame.
Signals of strength
The question names a real resource commitment, not a theme · A stated horizon, typically five to twenty years · Decision-makers agree the question matters before work starts
Identify driving forces
The researched inventory of external forces bearing on the focal question, sorted into predetermined elements and uncertainties. This is the evidence base that separates scenario planning from guessing in quadrants.
Signals of strength
Forces drawn from research and outsider interviews, not just the room · Predetermined elements explicitly listed and agreed · Forces traced to their own drivers rather than taken at face value
Isolate critical uncertainties
The ranking of forces by importance and uncertainty to find the two or three whose resolution genuinely reshapes the answer to the focal question.
Signals of strength
Each critical uncertainty has plausible, divergent outcomes · The chosen uncertainties are independent of each other · High-importance, low-uncertainty items routed into base planning instead
Build scenario logics
Two to four internally consistent narratives of how the future unfolds, each with a causal story, a memorable name and real consequences for the focal question.
Signals of strength
Every scenario is plausible to a sceptical executive · Scenarios differ in structure, not just in degree of optimism · No scenario is presented as the most likely
Rehearse implications and indicators
The payoff step: testing strategy in each world, extracting robust and contingent moves, and defining the early-warning indicators that connect the scenarios to ongoing monitoring and review.
Signals of strength
Current strategy explicitly stress-tested in each scenario · A short list of robust moves and real options recorded · Named indicators with owners and a review cadence
When it earns its keep
- The decisions in front of you have long payback periods, ten-year leases, plant, franchises, and the environment they depend on is genuinely uncertain rather than merely unmeasured.
- The organisation runs on a single official forecast, everyone privately doubts it, and planning has become the ritual of extrapolating last year with a growth percentage.
- An industry's fundamentals are in motion, through regulation, technology or shifting demand, and you need executives to take seriously futures they find uncomfortable.
- You want early-warning value: a shared set of named futures and indicators lets the organisation recognise which world it is entering while there is still time to act.
And when it doesn't
- The uncertainty is quantifiable risk with good data, demand seasonality, credit losses, where probabilistic forecasting and sensitivity analysis are sharper tools.
- You need a decision this week. A scenario exercise done in an afternoon produces labelled quadrants, not re-perception; a premortem or risk matrix gives faster value.
- Leadership wants the exercise to validate a decision already taken. Scenarios built to flatter a strategy are theatre, and expensive theatre.
- The organisation cannot act on divergent futures anyway, because capital, contracts or regulation lock it to one path. Rehearsing futures you cannot respond to demoralises rather than prepares.
How to run it
Before starting, gather the inputs the analysis depends on:
- A sharply framed focal question with a stated time horizon, tied to a real decision the organisation must make.
- A researched view of the driving forces in the environment: political, economic, social, technological, environmental and legal pressures bearing on the focal question, with evidence rather than opinion.
- A separation of predetermined elements, forces already in motion whose outcome is near-certain over the horizon, from genuine uncertainties.
- Access to the actual decision-makers. Wack's central lesson at Shell was that scenarios only work when they reach and change the mental models of the managers who commit resources.
- Diverse participants, including informed outsiders, to challenge the industry's consensus view of itself.
- 1
Frame the focal question
Anchor the exercise to a specific decision and horizon: should we commit to X over the next ten years? Unanchored scenario work drifts into futurism. The question determines which forces matter and stops the exercise sprawling.
- 2
Identify driving forces
Research the forces shaping the environment of the focal question. Distinguish predetermined elements, demographics already born, capacity already built, from uncertainties. Wack insisted this research phase is where the real work lies; scenarios assembled from a brainstorm alone inherit the room's existing assumptions.
- 3
Isolate critical uncertainties
Rank the forces by importance to the focal question and by uncertainty of outcome. The one or two that score high on both are the critical uncertainties. In the Schwartz and Global Business Network practice, two are crossed as axes to yield four candidate futures; treat that 2x2 as scaffolding, not as the deliverable.
- 4
Build scenario logics
Develop two to four scenarios into internally consistent narratives with causal logic: how the world gets there, who wins, what it feels like to operate in. Each must be plausible and structurally different, and none should be labelled the expected case, or everyone will plan for that one and file the rest.
- 5
Rehearse implications and indicators
Walk the current strategy through each scenario and record where it thrives, survives or dies. Extract robust moves that work across scenarios, options worth keeping open, and leading indicators that would signal which future is arriving. Wire the indicators into ongoing review so the scenarios stay alive after the workshop.
Reading the result
A set of two to four named, internally consistent scenarios; an assessment of how current strategy fares in each; a distinction between robust moves, contingent options and bets; and a monitored list of leading indicators tied to review cycles.
- Scenarios are rehearsal spaces, not forecasts. The question to ask of each is 'what would we wish we had done?', never 'which one do we believe?'.
- Weight the moves that perform across all scenarios; treat moves that only work in one future as options to price, not plans to fund.
- The exercise has worked when decision-makers describe the present differently afterwards. If the strategy and its assumptions emerge untouched, the scenarios were probably built around the official future.
A worked example
An agricultural machinery dealership group looks ten years out
A family-owned dealership group with twelve depots across East Anglia and the East Midlands holds franchises for a major tractor marque. Freehold depots need reinvestment, the franchise agreement is up for renewal, and the board is split between expanding metal sales and pivoting towards service. They commission a scenario exercise on the question: what should this group be by 2036?
- Frame the focal question
- The board sharpens the question to a decision: over the next three years we must commit roughly £8m across depot reinvestment, franchise renewal and technician recruitment; what mix survives the plausible futures of UK arable farming to 2036? Predecided items, one depot lease already expiring, are excluded.
- Identify driving forces
- Research and interviews with growers, the marque's UK arm and an agritech academic surface the forces: post-BPS subsidy design and environmental land management, farm consolidation, machinery electrification and autonomy, right-to-repair pressure, technician scarcity, and OEM interest in agency-style direct sales. Predetermined elements include the aging farmer population and the existing fleet's ten-year service tail, which guarantees workshop demand into the early 2030s whatever else happens.
- Isolate critical uncertainties
- Ranking leaves two forces both critical and genuinely open: the pace of autonomous and electric machinery adoption on UK arable farms, and whether OEMs shift to direct or agency distribution that demotes dealers to service agents. Subsidy design matters but converges on 'less money, more environmental strings' in every telling, so it is treated as predetermined.
- Build scenario logics
- Crossing the two uncertainties yields four logics, three of which survive plausibility testing: Iron Age Continues (slow autonomy, dealers keep the franchise relationship), Robot Fields (fast autonomy sold as per-hectare service, OEMs direct, dealers reduced to maintenance contractors), and Patchwork (fast autonomy on consolidated estates, conventional kit elsewhere, dealers straddling both). Each is written as a one-page narrative with a named grower archetype.
- Rehearse implications and indicators
- The expansion-led strategy thrives only in Iron Age Continues and is wounded in both others. Robust across all three: technician capability in software and electrics, service contracts sold on uptime, and depots configured for workshop throughput rather than showroom frontage. Indicators adopted: OEM agency-model announcements in any European market, autonomous hectares worked in the UK, and used-kit residual values. Reviewed at each half-year board meeting.
The read. The board declines the pure expansion path it had been leaning towards, renews the franchise but negotiates weaker exclusivity, and redirects £3m of the depot budget into workshop capacity and a technician academy. Nobody claims to know which future arrives. The honest reading is that the exercise bought optionality and a watchlist, and that its value now depends entirely on the board actually revisiting the indicators, which is where most scenario projects quietly die.
Pitfalls
- Building scenarios around the official future plus a good year and a bad year. Best, base and worst cases on one variable are sensitivity analysis wearing a costume.
- Producing four quadrant labels in an afternoon and calling them scenarios. Without the research phase and causal logic, the exercise recycles the room's assumptions with better graphics.
- Letting participants vote on the most likely scenario. The moment one future becomes the favourite, the others stop doing any work.
- Ending at the narratives. Scenarios without implications, options and indicators are corporate science fiction; the decision linkage is the whole point.
- Treating the scenario set as finished. Wack's Shell practice treated scenarios as a continuing conversation with management, updated as the world moved, not a bound report.
What the critics say
The dominant intuitive logics method rests on surprisingly thin evaluation evidence. Wright, Bradfield and Cairns conclude that claims for its effectiveness in changing organisational thinking and improving decisions are largely asserted rather than demonstrated, with few controlled studies of whether scenario interventions achieve their stated objectives.
Wright, G., Bradfield, R. and Cairns, G. (2013) 'Does the intuitive logics method, and its recent enhancements, produce "effective" scenarios?', Technological Forecasting and Social Change, 80(4), pp. 631–642.
The field's foundational success stories are vulnerable to hindsight selection. The Shell oil-shock episode is retold constantly while unsuccessful or inert scenario projects go unreported, and documented cases exist of interventions that entrenched rather than shifted managerial thinking.
Hodgkinson, G. P. and Wright, G. (2002) 'Confronting Strategic Inertia in a Top Management Team: Learning from Failure', Organization Studies, 23(6), pp. 949–977.
The method's own house is contested: reviews of the field find dozens of competing schools and techniques with inconsistent terminology and little agreement on what a valid scenario process is, which makes quality hard to judge and charlatanry hard to detect.
Bradfield, R., Wright, G., Burt, G., Cairns, G. and van der Heijden, K. (2005) 'The origins and evolution of scenario techniques in long range business planning', Futures, 37(8), pp. 795–812.
Run it as a workshop
This is a two-session exercise with a research gap between the sessions, not a single workshop, and should be billed to the client that way from the outset. Session 1 needs the actual decision-makers in the room, not delegates, since Wack's own lesson from Shell was that scenarios only change minds that were present when they were built. Before Session 1, brief the group on the focal decision and horizon so the framing conversation starts from a real commitment rather than a theme. Between the sessions, commission the driving-forces research properly: desk research plus a handful of outsider interviews, not a huddle producing a list from memory.
Running order
| Session 1 (half day): frame the focal question and agree the decision it must serve | 60 min |
| Session 1: brainstorm candidate driving forces and agree the research brief | 90 min |
| Session 1: assign research owners and the return date for Session 2 | 30 min |
| Session 2 (full day, held after the research gap): present driving-forces research and separate predetermined elements from genuine uncertainties | 90 min |
| Session 2: rank forces and isolate the critical uncertainties | 60 min |
| Session 2: build two to four scenario logics in small groups, then cross-challenge for plausibility | 120 min |
| Session 2: rehearse current strategy against each scenario and extract robust moves | 90 min |
| Session 2: agree named indicators, owners and the review cadence | 45 min |
| Total | 9h 45m |
You will need
- The focal question and decision brief, agreed and written down before Session 1 starts
- Large paper or whiteboards for driving-forces brainstorming and, later, the scenario logic maps
- The commissioned driving-forces research pack, circulated to all Session 2 attendees at least two days beforehand
- A simple 2x2 grid as scaffolding only, not the deliverable, for isolating the critical uncertainties
- A one-page narrative template per scenario, and an indicators-and-owners tracking sheet carried into ongoing review after the workshop
Traps to avoid
- Compressing this into one afternoon. A same-day scenario exercise produces labelled quadrants, not re-perception; the entry's own worked example calls this 'a costume' for sensitivity analysis, and it is the single most common way this framework gets run badly.
- Skipping the research gap, or filling it with the same room's opinions instead of genuine outside evidence. Wack was explicit that this phase is where the real work happens; a workshop that goes straight from brainstorm to scenario-writing inherits the room's existing assumptions.
- Letting the group vote on which scenario is most likely. The instant one future becomes the favourite, the room quietly stops planning for the others, and the exercise has failed at the thing it exists to prevent.
- Ending Session 2 at the narratives. A set of well-written scenarios with no rehearsal, no robust moves and no indicators is corporate fiction; budget real time for the implications step and do not let it be the first thing cut when the day overruns.
Sources and further reading
- Wack, P. (1985) 'Scenarios: Uncharted Waters Ahead', Harvard Business Review, 63(5), September–October 1985. ↗
- Wack, P. (1985) 'Scenarios: Shooting the Rapids', Harvard Business Review, 63(6), November–December 1985. ↗
- Schwartz, P. (1991) The Art of the Long View: Planning for the Future in an Uncertain World. New York: Doubleday.
- Bradfield, R., Wright, G., Burt, G., Cairns, G. and van der Heijden, K. (2005) 'The origins and evolution of scenario techniques in long range business planning', Futures, 37(8), pp. 795–812. ↗