Scenario planning is a structured method for building a small set of plausible, distinct futures around one strategic question, then testing your strategy against each of them. The output is not a prediction. It is robustness: knowing which moves work in every future you can defend, which depend on one future arriving, and which observable signals would tell you early which future is actually showing up.
The method earned its reputation at Shell in the 1970s, when scenario work left the company less surprised than its competitors by the oil shock. It has been re-explained many times since, usually in the abstract. This post walks it once, concretely, on a single running example.
The method, step by step
The example: a mid-market B2B software company, EUR 15M revenue, direct sales only, deciding whether to build a partner channel by 2028.
1. Sharpen the focal question
Scenario planning starts with a decision, not a topic. "The future of our market" produces essays. "Should we build a partner channel by 2028, and how aggressively?" produces scenarios. The question fixes the horizon (2028) and the stakes (a multi-year investment that is hard to reverse).
2. List the driving forces
The forces that could shape the answer, gathered wide before being filtered. For our company: consolidation among the system integrators it would partner with, AI eroding the implementation services partners sell, buyer preference shifting between best-of-breed and suite purchasing, its own category's pricing pressure, EU procurement regulation, the founder-led sales motion reaching its scaling limit.
Each force gets two ratings: impact on the focal question, and certainty about its direction. That grid does the filtering.
3. Pick the two critical uncertainties
High-certainty forces, whatever their impact, become givens inside every scenario. The founder-led motion hitting its limit by 2028 is near-certain, so every scenario contains it. The two axes must be high-impact and genuinely uncertain. Our company picks: whether implementation services remain a viable partner business under AI pressure (thriving vs hollowed out), and whether buyers consolidate purchasing into suites (best-of-breed holds vs suite wins).
4. Build the four scenarios
Crossing the axes gives four futures, each needing a name and a short narrative, because scenarios that read like spreadsheet rows never get used in an argument.
- Partner land grab (services thrive, best-of-breed holds): integrators actively seek software vendors to differentiate; channel economics are excellent; the risk is moving too slowly while competitors sign the good partners.
- Suite squeeze (services thrive, suites win): partners exist but orbit the big platforms; a channel is possible only as a niche play inside suite ecosystems.
- Direct in the dark (services hollowed out, best-of-breed holds): partners disappear as a route to market; the company must scale direct sales plus product-led motion; channel investment would be stranded.
- Platform gravity (services hollowed out, suites win): distribution collapses toward marketplaces; the channel question becomes a marketplace question.
5. Attach implications and options
For each scenario: what it would mean, which options it opens, which it closes. Cross-scenario reading is where the value concentrates. Two moves survive all four futures for our company: productizing onboarding so it depends less on partner services either way, and building marketplace presence. One move, an exclusive multi-year deal with a single large integrator, works brilliantly in one future and is a write-off in two. The exercise just converted a binary bet into a portfolio with triggers.
6. Define early-warning indicators
Each scenario gets observable signals with thresholds and owners: partner M&A volume, the share of deals where a suite was the finalist, services attach rates in industry reports. The indicators turn the scenario file from an offsite artifact into an instrument. When one crosses its threshold, the affected options get revisited within the month, not at next year's strategy day.
The scenario planning template in our library carries all six steps as a fill-in deck, free to download.
Strategic foresight: the discipline around the method
Scenario planning is one instrument inside a larger practice. Strategic foresight is the umbrella: horizon scanning for weak signals, trend analysis, Delphi panels that structure expert disagreement, and scenario work as the synthesis step. Governments and institutions run mature versions of this; the OECD and several national foresight units publish methods and studies that are genuinely worth reading, and horizon planning offers a lighter-weight framing for companies that find the full apparatus heavy.
Corporate foresight, though, has a characteristic failure that the institutional literature rarely names: the last mile. The scenarios get built, the report is sharp, the offsite goes well, and nothing in the operating company changes, because the scenarios never touch the actual work. The futures were described at the level of markets and forces; the organization runs at the level of goals, teams, and allocated hours; and no one translated between the two. Strategy already decays quickly on its own, and foresight that lives in a PDF decays fastest of all. The related failure runs in the other direction too: the organization pivots in response to events and the strategy documents never hear about it, the pattern we described in the silent pivot.
The fix is to ground the exercise in current reality. A scenario built on a live model of your goals and where capacity is actually allocated starts from what teams are doing today, so "what would we shift in the suite-squeeze future" is a concrete reallocation question with names on it, not a hypothetical. That live model of strategy connected to work is what we call a digital twin of strategy execution: the same twin you use to observe drift is the surface you test futures against.
Scenario planning vs forecasting
A forecast commits to one expected future, usually with a confidence interval, and works well where history is a guide: capacity planning, cash flow, seasonal demand. Scenario planning deliberately refuses to commit, and earns its overhead where the uncertainties are structural, the horizon is long, and betting on one future is expensive to reverse. The two are complements. Forecast what is forecastable; build scenarios around what is not. Treating a genuine structural uncertainty as a forecasting problem produces precise, confident, wrong numbers.
Scenario planning software
Honest answer: to start, a template is enough. The method is thinking work, and the deck linked above holds all of it. Dedicated foresight platforms add value at institutional scale, managing signal libraries and large scenario portfolios, but they share the last-mile problem: they model the outside world, not your execution.
The tooling gap that matters sits on the inside. When your goals, KPIs, and the work behind them live in one connected structure, scenario implications become testable against reality: which teams would a reallocation touch, which current goals would lose their capacity, which KPIs lose pace first. Vindaris holds that structure, so the scenario conversation ends with a simulated move on live data rather than a spreadsheet guess.
Keeping it alive
The failure rate of scenario work is mostly a maintenance failure. Three practices keep the exercise earning its cost: review the early-warning indicators inside an existing cadence rather than a separate ritual, revisit options within a month whenever an indicator crosses its threshold, and re-run the full exercise when an axis resolves or the focal question changes. A scenario file nobody re-reads was an offsite. The method deserves better, and so did the afternoon you spent on it.