Scenario Planning Step by Step: How Shell's Method Works

Most plans quietly assume one future: roughly today's world, plus a bit of growth. Scenario planning is a disciplined way to drop that assumption. Instead of betting everything on a single forecast, you build a small set of distinct, plausible futures and ask how your organization would fare in each.

Shell is the company most associated with the practice. It says it has been developing scenarios since the early 1970s, and in 2008 it published Scenarios: An Explorer's Guide, a book-length description of how it builds and uses them. This guide walks through that approach in plain English, with practical additions from the UK Government Office for Science's Futures Toolkit, so that a smaller team can run a lighter version.

What a scenario is, and what it is not

The Explorer's Guide puts it simply: "A scenario is a story that describes a possible future." Scenarios come in sets. Every scenario in a set addresses the same important questions and includes what Shell calls predetermined elements, the aspects of the future that are likely to persist. What differs is how the uncertain aspects play out.

Two points trip people up. First, scenarios are not predictions. Shell's own scenarios page describes them as "possible worlds built from incomplete and uncertain information." Second, they are not a consensus view. The guide notes that, unlike forecasting, scenarios do not demand consensus; they accommodate differences and try to define them clearly. It even argues that probability is not helpful when developing the scenarios themselves, because the real challenges often lie in an improbable turn of events.

Step 1: Prepare the project

Shell's first phase, Preparation, is about writing down what the project is for before anyone starts imagining futures. The guide suggests answering questions such as:

• What is the primary purpose: building knowledge, or improving how a team thinks and talks together?
• Who will use the scenarios: operational staff or the organization's top decision makers?
• What time horizon will the scenarios cover?
• How long do you have, who will be involved, and what will it cost?
• How will the scenarios be applied once they exist?

On the time horizon, the guide warns that if it is too short, scenarios may only convey a creative description of the present; if it is too long, they may lose focus and relevance. The Futures Toolkit, written for government policy work, suggests at least five years out and generally 10 to 20. A business may choose a shorter horizon, but the principle holds: look far enough ahead that the important things can genuinely change.

The main cost is usually people's time. Shell notes that workshop time on a large project can run to ten days or more. A small organization will scale that down dramatically, but it should still protect real time for the work.

Step 2: Research, then find themes and critical uncertainties

Shell calls the next phase Pioneering. The team gathers research across disciplines, deliberately reaching beyond its own expertise to expose blind spots, then merges that material into a smaller set of themes. For each theme, the guide's questions are: What forces drive it, and how do they relate? Which of those forces are relatively certain and which are most uncertain? For each uncertainty, what are the extreme outcomes, and which are most challenging and relevant?

The useful outputs are two lists: predetermined elements (things you are fairly sure of) and critical uncertainties (things that matter a lot but could go either way). If you want a structured way to list the outside forces, a PESTLE scan works well at this stage.

Step 3: Build the scenario structure

The Explorer's Guide describes a scenario structure in three parts:

• A focal question: the major challenge the users of the scenarios are likely to face.
• Branches: the different directions a critical uncertainty could take, each giving a different answer to the focal question.
• Scenario outlines: the stories created by following particular paths through the branches.

It offers three ways in. Deductive: pick two critical uncertainties, place the extremes of each on a matrix, and develop a storyline for each quadrant. Inductive: start from several chains of events (no fewer than three events in each) and work out what structure they imply. Normative: start from an end state you want to describe and work backward to see what it would take to get there and whether it is plausible.

The deductive 2x2 is the most common starting point, and it is the method the Futures Toolkit walks through. Its advice is to choose two axes that are independent of each other, so the four quadrants are genuinely different, and to name the scenarios carefully. The toolkit specifically warns against giving a positive name to the scenario you prefer and a negative name to the one you do not.

Step 4: Write and test the stories

In Shell's Map-making phase, an editor turns the material into readable narratives, and drafts circulate mainly to check internal consistency. The finished stories are then tested for plausibility, challenge and relevance. The guide cautions against excessive detail: precise events and timings can seem misleadingly like predictions. A scenario should give readers enough to imagine being in that world and to think about how they would behave in it.

One practical test from the guide: if people who did not help build the scenarios find them hard to understand, find out why. It might be presentation, or it might be a flaw in the logic of the story itself.

Step 5: Use the scenarios to test strategy

Scenarios earn their keep when they are used. The Explorer's Guide describes a simple workshop. Split participants into small groups, give each group one scenario, and ask each to list strategic options that would work well in its world. Then bring everyone back together and examine how each option plays out in every scenario.

Options that work well across all the scenarios are what Shell calls robust, and they are likely candidates for action now. Others are contingent on events. Some can be adjusted to become more robust, and some can be broken into stages: follow a strategy up to a certain point, then hold back until the direction of events is clearer.

After that, the scenarios become a lens for watching the world. The guide suggests scanning for signals that the dynamics behind one scenario are actually developing, so decision makers can begin to anticipate events rather than react to them.

A lighter version for a small team

For example, imagine a hypothetical 30-person regional distributor that is uneasy about the next decade. A lighter version might look like this:

1. Agree on a focal question, such as how the business will win and keep customers over the next ten years.
2. Spend two or three weeks collecting forces of change, then sort what is fairly certain from what is critically uncertain.
3. Pick two independent critical uncertainties and draw a 2x2 matrix.
4. In a workshop, describe the four worlds, name them neutrally, and list what each would demand of the business.
5. Test current plans against all four and agree which moves are robust.

Even a light version deserves a proper block of time. The Futures Toolkit suggests scenario building needs a minimum of 8 to 12 people and at least four to five hours of workshop time.

Finally, plan to refresh the work. Shell's guide observes that scenarios looking out over 20 years are likely to remain useful for 3 to 4 years, after which both the world and the organization have moved on enough to start a new cycle.

Key takeaways

• A scenario is a story about a possible future; a set of scenarios explores different ways the key uncertainties could play out.
• Scenarios are neither predictions nor consensus forecasts, and probabilities are not the point when building them.
• Shell's approach moves from preparation through research, structure and storytelling to application.
• A 2x2 matrix built on two independent critical uncertainties is the simplest structure; name the quadrants neutrally.
• The payoff is testing strategy: look for options that hold up in every scenario, and stage the rest.
• Refresh scenarios periodically as the world and your organization change.

Sources

• Shell International, Scenarios: An Explorer's Guide (2008)
• Shell, What are Shell Scenarios?
• UK Government Office for Science, Futures Toolkit: tools for strategic futures for policymakers and analysts (2024 edition)