A Strategy That Only Works in One Future Is a Bet, Not a Plan
Published 2026-10-02 · EuroQuest International
Quick summary
- It is not forecasting. A forecast picks the most likely future. Scenario planning builds several plausible ones on purpose, so the strategy is tested against the futures nobody in the room expects.
- Do not assign probabilities. The moment one scenario is labeled "most likely", the room stops taking the others seriously and the exercise collapses back into a forecast.
- Two uncertainties, four futures. Pick the two drivers that are both highly important and genuinely unpredictable, cross them, and the four corners give you a set that is wide without being unmanageable.
- The output is decisions, not documents. A scenario set that does not change at least one choice, hedge or trigger has been an interesting workshop rather than planning.
- Signposts make it live. Each scenario needs a few early indicators that would show it starting to arrive, and someone named to watch them.
Most strategies are built on a single picture of the future, usually an extrapolation of the last three years with some adjustments for whatever worried the leadership team most recently. The plan is then judged on whether it is a good response to that picture. What nobody asks, because the format does not invite it, is how the plan performs if the picture is wrong. Scenario planning exists to ask exactly that question, in a structured way, before events ask it for you.
This guide covers what scenario planning is and what it is not, how to frame the question so the work stays useful, how to choose the drivers that genuinely matter, how to turn them into futures that a skeptical leadership team will engage with, how to test strategy against them, and how to set signposts so the work keeps paying after the workshop ends. It is a method guide rather than advice on any particular market. The public examples below come from central banks, climate science and fiscal policy because those are the institutions that publish their scenario work in full; the method itself applies to any organization making long-lived commitments under uncertainty.
On this page
- What is scenario planning, and what is it not?
- What question should the exercise answer?
- How do you choose the drivers that matter?
- How do you turn drivers into futures people take seriously?
- How do you test a strategy against them?
- What are signposts, and who watches them?
- Frequently asked questions
What Is Scenario Planning, and What Is It Not?
Scenario planning is a structured way of describing several distinct, internally consistent futures and then asking how a strategy performs in each. The futures are not predictions. They are chosen to be different from one another in the ways that matter most for the decision at hand, so that between them they span the range of outcomes the organization could plausibly face.
The idea is easiest to see in a published example. The Intergovernmental Panel on Climate Change, in its most recent assessment of the physical science, states that it "assesses the climate response to five illustrative scenarios that cover the range of possible future development of anthropogenic drivers of climate change found in the literature." The important word is range. The five were not chosen because one of them is expected; they were chosen because together they bracket what could happen.
It Is Not Forecasting
A forecast tries to say what will happen and attaches a confidence to it. A scenario set deliberately refuses to do that. The same IPCC summary is explicit on the point: "the feasibility or likelihood of individual scenarios are not part of the assessment." That restraint is not modesty. It is the mechanism that makes the method work, because as soon as one scenario is labeled most likely, every discussion gravitates to it and the others become decoration.
Organizations that already forecast well sometimes resist scenarios as vague. The answer is that the two tools do different jobs. Forecasting is the right instrument for the next budget year, where the range of outcomes is narrow and history is a fair guide. Scenarios are the right instrument for commitments that outlive the forecast horizon: a plant, an acquisition, a market entry, a ten-year contract, a workforce built around one technology. Those are the decisions where being confidently wrong is most expensive.
It Is Not a Tabletop Exercise or a Risk Assessment
Three neighbors get confused with it. A tabletop exercise, covered in the guide on how to run a tabletop exercise, tests how people respond to a single incident over hours or days; scenario planning tests a strategy against whole environments over years. A risk assessment, as set out in how to conduct a risk assessment, ranks specific known risks by likelihood and impact; scenario planning works with the forces that shape which risks exist at all, and deliberately avoids likelihood. And a SWOT analysis, explained in what a SWOT analysis is, describes the organization's position now. Scenarios describe the conditions it might be in later. Each feeds the others, and none replaces them.
What Question Should the Exercise Answer?
The most common reason scenario work fails is that it starts without a decision attached. A team is asked to imagine the future of the industry, produces four vivid stories, presents them to applause, and nothing changes. The scenarios were about the world in general, and the world in general is not something the organization decides about.
The Focal Question and the Horizon
Start with a focal question: the specific choice the scenarios will inform. Should we build the second facility? Should we commit to this technology platform? How much of the workforce should we reskill for automation, and when? A focal question keeps the work bounded, because it determines which forces are relevant and which are interesting but beside the point.
Then set the horizon by the life of the commitment rather than by habit. A decision whose consequences run for fifteen years needs scenarios that reach fifteen years, even though everything beyond five feels speculative. That discomfort is the point. Fiscal planners face it routinely: the US Congressional Budget Office's September 2026 report on alternative scenarios projects federal debt held by the public rising from 101 percent of GDP in 2026 to 175 percent in 2056 under its extended baseline, to 222 percent under a scenario with higher interest rates, and staying at 101 percent under a scenario in which the ratio is held constant. Same institution, same starting point, three answers thirty years out, each one coherent. That is what a properly framed horizon produces, and the spread is the information. Framing questions this way is the core of future trends in leadership and strategic planning work.
How Do You Choose the Drivers That Matter?
Drivers are the forces that will shape the environment the decision plays out in: regulation, technology, customer behavior, input prices, interest rates, demographics, geopolitics, competitor moves. A workshop will happily generate sixty of them. The skill is in sorting, not in listing.
Separate What Is Fixed From What Is Genuinely Uncertain
Sort every driver on two questions: how much does it matter to the focal question, and how uncertain is it? Some important drivers are not uncertain at all. Population aging in most developed economies, for instance, is largely locked in for the next two decades because the people involved are already born. These are predetermined elements, and they belong in every scenario. Treating them as uncertain wastes the exercise on false variety.
What remains in the top right of the grid, high importance and high uncertainty, are the critical uncertainties. These are the forces that will genuinely differ between futures and genuinely change the right answer. Most teams find between four and eight. The discipline is in being honest about which are truly uncertain rather than merely uncomfortable, and in resisting the urge to promote a favorite concern into the top tier. Structured judgment of this kind is the subject of strategic thinking and decision making.
The Two-Axis Method, and Why It Works
The most widely used construction takes the two critical uncertainties that are most important and least correlated with each other, treats each as an axis running between two extremes, and crosses them. The four quadrants become four scenarios. It sounds mechanical, and the mechanics are the virtue: the result is guaranteed to be wide, internally consistent, and different along the dimensions that matter most.
The choice of axes is where most of the thinking happens. Two axes that move together, such as economic growth and consumer demand, produce two near-identical scenarios and two implausible ones. Two axes that are independent, such as the speed of regulation and the cost of a key technology, produce four genuinely distinct worlds. Test candidate pairs quickly, sketch the four corners in a sentence each, and keep the pair that yields four futures you can each imagine a competent observer defending.
How Do You Turn Drivers Into Futures People Take Seriously?
A grid of four quadrants is a structure, not yet a set of scenarios. The work of making it persuasive is narrative, and it is where scenario programs either earn the leadership team's attention or lose it.
Name Them, and Write Them as Stories With a Path
Give each scenario a short, neutral, memorable name, and write it as a plausible chain of events from today to the horizon rather than as an end state. A story with a path is far harder to dismiss than a description of a destination, because the reader can see each step and has to argue with a specific step rather than with the whole idea.
The Bank of England's 2021 climate exercise, whose results were published in May 2022, is a clean public example of naming. Banks and insurers were tested against three scenarios: "Two scenarios featured policies to limit global temperature rises (Early Action scenario and Late Action scenario), the third featured unchecked global warming (No Additional Action scenario)." The names describe what differs, carry no judgment about which is better, and are short enough to use in conversation. The same release is candid about the limit of the exercise: "Projections of climate losses are uncertain." Scenario work that says so plainly is more credible, not less.
Three or Four, and Never a Base Case
Two scenarios turn into an optimistic and a pessimistic case, and everyone plans for the middle. Five or more become too many to hold in mind or to test strategy against properly. Three or four is the working range. Avoid an odd-numbered set built around a central scenario, because the middle one becomes the de facto forecast and the outer two become the margins of error around it.
The same applies to language. Do not call any scenario the base case, the expected case or the reference case inside the exercise. Those labels are sometimes necessary for external reporting, but in the room they reintroduce the forecast the method is designed to suspend. The central banks behind the shared climate scenarios published by the NGFS, a network that brought together 141 central banks and supervisors and 21 observers when it released its fifth long-term set in November 2024, describe the work as exploring impacts "over a long time horizon and under varying assumptions." Varying assumptions is the right frame, and the managing of genuine complexity it implies is covered in managing complexity and strategic execution.
How Do You Test a Strategy Against Them?
This is the step that turns a scenario set into planning, and it is the one most often cut for time. The method is usually called wind-tunneling: take the current strategy, or each strategic option under consideration, and run it through every scenario in turn.
Ask the Same Four Questions in Every Future
For each scenario, ask whether the strategy still works, what would break first, what you would wish you had done differently with five years' hindsight, and what you could do now that would help in this future without costing much in the others. The answers sort options into three groups. Some moves perform acceptably in every scenario; those are the no-regret moves and they should be made regardless. Some moves pay off handsomely in one or two futures and badly in others; those are bets, and they should be made deliberately and sized accordingly. And some moves are cheap ways to keep an option open; those are hedges.
| Type of move | How it behaves across scenarios | What to do with it |
|---|---|---|
| No-regret move | Acceptable or better in every scenario | Make it now; the exercise has just removed the argument against it |
| Bet | Strong in one or two scenarios, poor in the others | Make it only deliberately, sized to what you can afford to lose, with a named signpost that would confirm or kill it |
| Hedge | Costs a little in every scenario, pays off in one | Buy it if the payoff scenario would otherwise be severe; an option, a pilot, a supplier relationship kept warm |
| Trap | Looks attractive in the expected future, fails badly elsewhere | Drop it, or redesign it so its failure in the other scenarios is survivable |
| Trigger-based move | Right in some scenarios, wrong in others, and you can tell which in time | Pre-agree the action and the signpost that triggers it, so nobody has to win the argument again later |
The trap row is the one that justifies the whole exercise. Most organizations have at least one strategic commitment that is excellent if the future resembles the past and disastrous if it does not, and nobody has noticed because the planning process only ever looked at one future. Finding it is worth a year of workshops. Weighing which risks to carry and which to shed is the territory of risk management in uncertain markets.
What Are Signposts, and Who Watches Them?
A scenario set that is filed after the workshop decays within months, because the organization stops looking for the evidence that would show which future is arriving. Signposts are what keep it alive.
Choose Indicators That Move Early
For each scenario, identify two or three observable indicators that would move early if that future were starting to arrive: a regulatory consultation opening, a price crossing a threshold, a competitor making a specific commitment, a customer segment changing behavior in a measurable way. Good signposts are specific, observable without special access, and early enough to act on. A signpost that only moves once the scenario is obvious is a headline, not a warning.
Assign each signpost to a named person, review the set on a fixed rhythm, quarterly for most organizations, and connect each to a pre-agreed response where one exists. That last link is what converts the trigger-based moves in the table above into actual decisions. Without it, a signpost moves, someone notices, a meeting is called, and the argument the exercise was supposed to settle starts again from the beginning. Acting quickly once a signpost moves is its own skill, taught as crisis decision making for business leaders.
Refresh the Set, Do Not Rebuild It
Revisit the scenarios annually and after any major surprise. Most years the right action is to update the narratives and signposts rather than start again, because the critical uncertainties usually stay critical for longer than people expect. Rebuild only when a critical uncertainty has resolved, at which point it becomes a predetermined element and a new uncertainty takes its place. Fresh drivers increasingly come from data and modeling as well as judgment, which is where AI and data-driven strategy earns a place in the process.
Where Teams Build This Capability
EuroQuest International runs strategy and foresight programs for leadership teams, strategy functions and planning staff in Geneva, Madrid, Zurich, Amman and Istanbul. Sessions are built around a real focal question the participants bring, so the group leaves with a working scenario set and a signpost list rather than a method on paper. The full range sits under leadership and strategic management.
Frequently Asked Questions
What is the difference between scenario planning and forecasting?
A forecast tries to say what will most likely happen and attaches a confidence to it. Scenario planning deliberately builds several plausible futures and refuses to say which is most likely, so that a strategy can be tested against all of them. The two do different jobs. Forecasting suits short horizons where history is a fair guide and the range of outcomes is narrow, such as next year's budget. Scenarios suit commitments that outlive the forecast horizon, such as a plant, an acquisition or a market entry, where being confidently wrong is most expensive. Most organizations need both, used for the decisions each is suited to.
How many scenarios should you build?
Three or four. Two tend to collapse into an optimistic and a pessimistic case, and everyone quietly plans for the midpoint. Five or more become too many to hold in mind or to test a strategy against properly. Four is the natural number when the set is built by crossing two critical uncertainties, which is the most common method. If you build three, avoid making one of them a central case flanked by two extremes, because the middle one becomes the forecast the exercise was meant to suspend, and the outer two become margins of error around it rather than genuinely different worlds.
Should you assign probabilities to scenarios?
No, not inside the exercise. As soon as one scenario is labeled most likely, discussion gravitates to it and the others stop being taken seriously, which turns the work back into forecasting. The IPCC's assessment of the physical science makes the same choice explicitly, stating that the feasibility or likelihood of individual scenarios are not part of its assessment. If a separate process genuinely needs probabilities, for financial reporting or capital modeling for instance, produce them in that process and keep them out of the strategy discussion where they would distort it.
How long does a scenario planning exercise take?
A useful first set can be built in two or three structured workshops over a few weeks: one to frame the focal question and sort the drivers, one to build and name the scenarios, and one to test strategy against them and set signposts. Preparation between sessions matters as much as the sessions, particularly gathering evidence on the drivers so the debate is about the world rather than about opinions. Large organizations with a dedicated strategy function sometimes run a longer program with external interviews. Either way, the ongoing cost after the first set is small: a quarterly signpost review and an annual refresh.
How do you know the exercise worked?
By whether it changed a decision. A scenario set that leaves every existing plan untouched has been an interesting workshop rather than planning. Useful signs are that at least one no-regret move was made, at least one strategic commitment was redesigned because it failed badly in one of the futures, at least one hedge or option was bought, and signposts with named owners are being reviewed on a fixed rhythm. A further sign appears later: when an unexpected development arrives, the leadership team recognizes it as the early stage of a scenario it has already discussed, and responds faster because the argument has already been had.
Test your strategy against more than one future
EuroQuest International runs practitioner training in strategic planning, foresight, decision making and leadership under uncertainty across Europe, the Gulf and Asia.