ProcessGeopolitics
Geopolitical scenario planning, from focal question to trigger points
Geopolitical scenario planning earns its cost when it is built around one decision and ends in actions with triggers. This method takes a team from a focal question and time horizon through driving forces, critical uncertainties and internally consistent scenarios, then uses those scenarios to stress-test the plan and sort responses into no-regret moves, hedges, real options and trigger points. It also explains when probability weights help and when they mislead.
On this page
- Start with the decision, not the region
- The scenario planning sequence at a glance
- Seven working steps and the output of each
- Choosing a scenario structure
- When probability weights help and when they create false precision
- Sorting responses into moves, hedges, options and triggers
- A hypothetical manufacturer facing tariffs and shipping disruption
- Questions and answers
- Sources
Start with the decision, not the region
The most common failure in geopolitical scenario work is starting from a region or a headline. The team produces vivid stories about how a rivalry or a conflict might unfold, the board finds them interesting, and nothing in the plan changes. Scenarios become useful when they answer a question someone has to decide: should we add a second assembly site outside our current country before the end of next year? That question names the decision, implies an owner and sets the horizon.
The horizon follows from the commitment. A supply contract renewal may need a view of the next two years; a plant with a long payback needs a decade or more. Choosing a horizon longer than the commitment invites speculation, while a shorter one hides the risks that will matter most.
ColdAI's geopolitics practice uses scenario development to stress-test client strategies and supply chains against each scenario and find vulnerabilities1. The sequence below is the working method. For a broader catalogue of foresight tools, the UK Government Office for Science's Futures Toolkit describes driver mapping, scenario axes and policy stress-testing in more depth2.
The scenario planning sequence at a glance
- Focal question
One decision, its owner and the horizon it must cover.
- Driving forces
Political, trade, security, technology, energy and regulatory forces with a path to the decision.
- Critical uncertainties
The few forces that are both high-impact and unresolved, separated from what is already in motion.
- Scenario narratives
Internally consistent stories of how each future could arise from today.
- Signposts and watchers
Observable indicators for each scenario and the person who tracks them.
- Stress test
The current plan run through every scenario to find where it breaks.
- Responses and triggers
No-regret moves, hedges, real options and pre-agreed trigger points.
Seven working steps and the output of each
Set the focal question and horizon
Write the decision as a question with a clear choice, record who will make it and by when, and set the horizon to match the life of the commitment.
List the driving forces
With people who know the regions and sectors involved, list political, trade, security, technology, energy and regulatory forces. For each, write how it could reach the decision: through cost, market access, legality, people or assets. Drop forces with no plausible path.
Separate predetermined elements from critical uncertainties
Predetermined elements are already in motion and will play out in every scenario, such as an enacted law with a fixed application date or a competitor's plant under construction. Rate the remaining forces on impact and uncertainty. The high-high group supplies the scenario axes, but only if the forces are independent: two uncertainties that tend to move together form one axis.
Build the scenarios
Cross two independent uncertainties into four scenarios, or select a few coherent combinations when three or more matter. Write each as a narrative of how the world gets there from today, with named actors and a plausible sequence of events, then check that nothing in it contradicts itself or the predetermined elements.
Assign signposts and watchers
For each scenario, list early, observable indicators that it is unfolding, such as legislative steps, official statements, trade investigations, shipping insurance terms or licensing announcements. Name the person or monitoring feed responsible for each, and define what change counts as a signal.
Stress-test the strategy
Run the current plan through every scenario. Ask what happens to revenue, supply, people, physical assets and legal position, where the plan breaks, and how long and how costly recovery would be.
Choose responses and set triggers
Sort candidate actions using the decision table below, write trigger rules into the plan with named owners, and agree when the scenario set will next be reviewed.
Choosing a scenario structure
| Criterion | Two-axis matrix | Multi-axis selection | Single-uncertainty range |
|---|---|---|---|
| Best when | Two independent uncertainties clearly dominate | Three or more uncertainties interact and all matter | One variable dominates, such as a tariff level |
| How many scenarios | Four, by construction | A few chosen combinations, usually three to five | Usually three: low, central and high |
| Main strength | Easy to explain and forces real contrast between futures | Captures interactions a matrix flattens | Fast, and maps straight onto financial models |
| Main weakness | Can hide a third uncertainty that matters | Choosing combinations takes judgement and can look arbitrary | Misses discontinuities and combined shocks |
| Typical use | Siting, market entry and board-level strategy | Supply networks spanning several jurisdictions | Budget sensitivity and pricing |
Start with the two-axis matrix unless the uncertainty ranking clearly shows more than two dominant forces; a single-uncertainty range is a sensitivity analysis rather than full scenario planning.
When probability weights help and when they create false precision
Probability weights help when there is a reasonable reference class, for example how often trade investigations of a particular kind have ended in duties, and when the decision is genuinely about expected value, such as allocating buffer stock across several inputs. In those cases, rough weights make trade-offs explicit and can be updated as signposts move.
They mislead when scenarios are deliberately divergent, when the events have little history, or when one number becomes the headline. A scenario labelled as the most likely quietly becomes the plan, and the purpose of the exercise, which is robustness across futures, is lost. Where weights are used, express them as bands or ranges, record who made the judgement and on what evidence, and keep likelihood separate from confidence.
Sorting responses into moves, hedges, options and triggers
- If
An action improves the plan in every scenario.
ThenTreat it as a no-regret move and do it now, for example qualifying a second supplier for a component already close to capacity.
Waiting for clarity adds nothing when the action pays off whichever future arrives.
- If
An action protects against one or two damaging scenarios at moderate cost.
ThenHedge: buy protection that limits the loss, such as extra stock of one input or contract clauses that allow re-routing.
A hedge caps the downside without committing to a view of which future is coming.
- If
An action would be valuable in some scenarios but is expensive to do in full now.
ThenBuy a real option: a small commitment that preserves the right to act later, such as an option on land, a pre-qualified contract manufacturer or a dormant legal entity.
Options keep the expensive move available at a fraction of its cost.
- If
A large, irreversible move only makes sense in one scenario.
ThenSet a trigger point: write down which signposts, at what level, would lead to the move and who has the authority to call it.
Agreeing the trigger in advance avoids both panic and paralysis when the signal arrives.
A hypothetical manufacturer facing tariffs and shipping disruption
Questions and answers
How many scenarios should a geopolitical scenario exercise produce?
Usually three to five. Four is common because crossing two independent uncertainties produces four distinct futures. Fewer than three tends to collapse into a base case with optimistic and pessimistic variants, while more than five becomes hard to stress-test properly and hard for decision-makers to remember. The right number is the smallest set that captures the futures in which your plan would perform very differently.
How often should geopolitical scenarios be refreshed?
Review them when a signpost moves materially, when the decision they support changes, and on a fixed cycle tied to planning, often annually. Scenarios built for a one-off decision can be retired once it is made and its triggers are handed to monitoring. Scenarios that support an ongoing strategy need their driving forces revisited, because new forces appear and old uncertainties get resolved.
Who should take part in a scenario workshop?
The decision owner, people from the functions that would feel the impact, such as operations, procurement, finance, legal and sales, and analysts who know the regions involved. Include at least one person likely to challenge the consensus. Keep the group small enough to argue productively, and make sure the person who will act on the result attends the stress-test session, not just the final readout.
Can AI tools build geopolitical scenarios?
They can help: language models are useful for scanning sources for driving forces, drafting first versions of narratives and tracking signposts across many documents. They should not decide which uncertainties matter or judge whether a narrative is plausible, and they can produce confident-sounding detail no source supports. Treat model output as a draft that analysts check against evidence and the decision at hand.
Sources
- Geopolitics capability: approach and offerings — ColdAI
- Futures toolkit for policymakers and analysts — Government Office for Science · checked 10 October 2026