Foresight and Scenario Planning

For most companies, the five-year plan is a single series of numbers: growth at this percent, margin at that, market share at this level. The plan sits on top of a forecast, and when the forecast fails, the plan collapses. Then management debates why the plan did not hold, when the real question is why it depended on one future so heavily.

The starting point of scenario planning is this: the future cannot be known, but the structure of the unknown can be thought through. Which uncertainties will truly shape our business, and are we ready for each?

When you need it

You need it if a few large external uncertainties decide your company's fate: regulatory change, technological breaks, energy and raw material prices, the currency and rate environment, customer behavior, geopolitical risk. If your investments have long payback periods (a plant, infrastructure, a large platform build, a new market), the price of uncertainty is large too.

The signals are concrete: the last plan's assumptions became invalid within six months, everyone on the management team seems to assume the same future, events labeled “unexpected” happen several times a year. It is too early if cash runs out in a few months. There the need is not a long-term scenario but a short-term cash scenario, and even a plain three-case cash table helps.

What most teams miss

Scenario planning often gets reduced to a forecast table split into “optimistic, base, pessimistic.” That is not scenarios, it is three scales of the same story. Real scenarios are qualitatively different worlds: one where regulation tightens and one where it loosens, one where a new technology spreads fast and one where it stalls. The difference between them lies not in the numbers but in the rules of the game.

The second common mistake is treating scenarios as a bet on which one is right. “Which will happen?” is the wrong question. The right one is: which of our decisions work in all scenarios, and which only in one?

Scenarios are written not to know the future, but to notice early which future is arriving and to move fast accordingly.

The best-known historical example is Shell's work in the 1970s. Pierre Wack and his team prepared scenarios that stepped outside the usual assumptions about oil supply and prices, and it is widely told that having discussed those scenarios with management beforehand made Shell's reaction faster during the 1973 oil shock. The lesson is not the accuracy of a prediction but that the managers' mental models had been loosened in advance.

The third thing missed is that scenarios fail to produce decisions. An impressive presentation is made, then its shelf life ends. Scenarios create value only when tied to investment decisions, strategic options and early warning signals.

The playbook

1. Set the focal question

“The future” is a broad subject; scenarios should serve a specific decision. Extract a concrete question such as “what position do we want in this market in ten years, and which investments should we make?” The output is a one-sentence focal question and a decision date.

2. Pick two high-impact, high-uncertainty drivers

List the external factors and rate each on “how large is the impact” and “how uncertain is it.” Choose the two drivers that are high on both, and draw a two-axis matrix. Each quadrant represents a world. Place the certain trends (such as demographics) in all scenarios as a base.

3. Write the four worlds

One story per quadrant: in that world, how do customers behave, what are competitors doing, what does regulation look like, how does our business look? Name each, keep it short, no more than a page. The leadership team does not have to agree on how plausible each scenario feels; that is not the point.

4. Test the current strategy in each world

How does your plan perform in each scenario? Which investments work in all worlds (robust decisions), which in only one (bets), which are a disaster in one world (risks)? Also look for ways to buy options at small cost: a pilot, a partnership, staged investment, a flexible contract.

5. Define early warning signals and triggers

For each scenario, pick three to five observable signs that it is materializing: a draft regulation, a technology crossing a cost threshold, a competitor move, a customer behavior indicator. Tie each signal to a trigger in the form “if this threshold is crossed, we make this decision.”

6. Attach it to a rhythm

Review scenarios once a year and the signal board every quarter. Scenario planning is not a one-time workshop but a habit that keeps management's view of the world alive.

What to measure

• Movement on the signal board: which scenario the monitored indicators are drifting toward; a leading indicator.
• Robust decision ratio: how many investments give an acceptable outcome across all scenarios.
• Trigger-to-decision time: how fast the company moves once a signal threshold is crossed.
• Option portfolio: the number of open options acquired at small cost that can be scaled later.

Pitfalls

Mistaking three forecasts for scenarios. Three versions of the same curve do not show different worlds. A scenario without a qualitative difference is only sensitivity analysis.

Producing too many scenarios. Six or seven scenarios do not live in management's memory. Two or three drivers and four worlds are enough.

Staying disconnected from decisions. If scenarios were reported and shelved, they did not work. Every exercise should change at least one investment or contract decision.

Putting everything on the most likely scenario. Scenarios are not a probability contest. Concentrating all resources on the world that looks most likely is itself a product of the wish to avoid uncertainty.

Start this week

1. Write down the ten key assumptions of your last strategic plan and note beside each: “what happens if this turns out wrong?”
2. Ask each member of the management team separately: “what are the two external uncertainties that could shake our business most in the next five years?” Lay the differences side by side.
3. Pick your largest upcoming investment decision and run a one-hour session on the question “in which world does this decision turn out wrong?”