What is scenario planning?
Scenario planning is the practice of building a small number of coherent, alternative views of the future — and working out, in advance, what each would do to your plan and what you would do about it. Instead of betting the business on one forecast, finance models three or four internally consistent futures (a base case plus plausible upsides and downsides), quantifies each through the operating plan, and pre-agrees the triggers and actions that each would demand. The output is not a prediction; it is preparedness. (Short version in our glossary entry.)
The discipline earns its keep on one insight: by the time a shock is confirmed, the best responses are usually gone. The company that worked out in March what a 10% currency move would do to gross margin — and which price rises, hedges and cost actions it would trigger — executes in days. The company meeting the question for the first time in the crisis spends its first month arguing about the model. Scenario planning buys speed, and speed is usually worth more than accuracy.
It also fixes a quieter problem: single-number plans manufacture false confidence. A board that has only ever seen the base case will treat any deviation as failure. A board that has seen the range treats deviation as navigation. (On why the base case itself is not a promise, see forecast vs projection vs budget.)
Where scenario planning fits in the planning cycle
Scenario planning is not a separate annual ritual; it attaches to the planning artefacts you already produce. During budget season, scenarios pressure-test the annual plan before the board approves it — a budget that survives the downside case is a materially stronger commitment than one that has only ever been viewed from the base case. Through the year, the scenario set rides along with the rolling forecast: each re-forecast updates the base case, and the downside and upside cases update with it, so the range around the outlook stays current rather than decaying into last autumn's assumptions.
The horizon should match the decisions at stake. For operating decisions — hiring gates, discretionary spend, pricing — twelve to eighteen months is right. For financing and capacity decisions, stretch the same scenarios to three years, at lower resolution. What matters is that all horizons run off one driver model, so a scenario means the same thing wherever it appears.
Scenario planning vs sensitivity analysis vs stress testing
Three related tools, routinely conflated, answering different questions:
| Sensitivity analysis | Scenario planning | Stress testing | |
|---|---|---|---|
| What varies | One variable at a time | Several variables together, as a coherent story | Variables pushed to extremes, often combined adversely |
| Question answered | Which assumptions matter most? | What might plausibly happen, and what would we do? | What breaks us, and where are the limits? |
| Output | A ranked list of value drivers (often a tornado chart) | 3–4 quantified futures with trigger points and playbooks | Survival thresholds: covenant breach points, cash runway floors |
| Typical use | Deciding which drivers deserve scenario treatment | Planning, board discussion, resource allocation | Financing decisions, covenant negotiation, risk committees |
They sequence naturally: sensitivity analysis tells you which two or three drivers move the outcome enough to matter; scenario planning combines those drivers into plausible futures; stress testing pushes the worst of them until something structural breaks. A team doing only one of the three is usually doing scenario planning badly — flexing one assumption ±10% and calling the result "scenarios".
A five-step scenario planning framework
- 1. Identify the critical uncertainties. Run sensitivities across your driver model and keep the variables that are both high-impact and genuinely uncertain — typically demand volume, pricing, a key input cost, FX, and the timing of a major initiative. Discard the merely interesting; three drivers explain most businesses.
- 2. Build three or four coherent scenarios. A base case, a plausible downside, a severe-but-plausible downside, and (optionally) an upside — because unpreparedness for good news is also expensive. Coherence is the discipline: in a recession scenario, volumes fall and debtors slow and pricing pressure rises together. Internally inconsistent scenarios are the most common technical failure.
- 3. Quantify through the driver model. Each scenario is a set of driver inputs run through the same model that produces the plan — never a hand-adjusted copy of the budget spreadsheet. If your model cannot re-run a full P&L, cash flow and covenant check from changed drivers in minutes, fix that first; it is the difference between scenario planning and scenario theatre.
- 4. Pre-agree triggers and actions. For each scenario, define the observable early indicators (order intake below X for two consecutive months; spot rate through Y) and the actions each trigger releases: hiring freeze tiers, discretionary spend gates, price actions, hedging, drawdown of facilities. This step converts analysis into an executable playbook — and it is the step most teams skip.
- 5. Monitor and refresh. Report the trigger indicators monthly alongside actuals, and refresh the scenario set quarterly or when the world lurches. A scenario deck from last January protects nobody.
A well-run first pass through these five steps takes a small team two to three weeks, most of it in steps 3 and 4 — and the elapsed time is dominated by the state of the driver model, not by the scenario thinking itself. Subsequent refreshes should take days. If each quarterly refresh still costs weeks, the model is the bottleneck and the highest-return fix is there, not in more sophisticated scenario narratives.
Worked example: an FX and demand shock
A UK distributor imports most of its product from US suppliers. Base plan: revenue £40.0m, cost of goods £26.0m (65% of revenue, of which 70% is US-dollar denominated), operating costs £10.0m — EBITDA £4.0m at a 10% margin.
The downside scenario combines two individually survivable shocks: demand softens 8%, and sterling weakens 10% against the dollar. Coherently, they arrive together — the same macro conditions drive both.
| Base plan | Downside scenario | |
|---|---|---|
| Revenue | £40.0m | £36.8m (volume −8%) |
| Cost of goods | £26.0m | £25.8m (volume-adjusted £23.9m; USD-denominated £16.7m revalued +11% to £18.6m) |
| Operating costs | £10.0m | £10.0m (largely fixed in-year) |
| EBITDA | £4.0m | £1.0m |
Two "modest" shocks — high single digits each — compound into a 75% EBITDA decline, because the FX move claws back almost all of the cost relief the volume drop would otherwise deliver. That non-obvious interaction is precisely what single-variable sensitivity analysis misses and scenario planning exists to expose.
The playbook that falls out: hedge a defined portion of the next two quarters' dollar purchases now, at base-case rates; pre-agree a supplier renegotiation and a targeted price increase that trigger if the spot rate holds below the threshold for a month; tier the discretionary cost base so £0.8m can be gated quickly; and test the scenario against banking covenants before it is needed, not after.
The cash dimension deserves its own line in any real exercise. The P&L damage above understates the strain: in the same scenario, customers pay more slowly while dollar-denominated suppliers still expect payment on terms — so working capital absorbs cash exactly when EBITDA stops generating it. Running the scenario through the cash flow and the covenant calculations, not just the P&L, is what turns an interesting analysis into one the CFO can act on.
Common scenario planning mistakes
- Scenarios that are just the budget ±10%. Scaling every line by the same factor produces a smaller or larger version of the same future, not a different one. Real scenarios change the shape of the P&L — mix, margins and cash conversion move differently, as the worked example shows.
- Only modelling downsides. An unplanned surge in demand strains working capital, capacity and hiring just as surely as a slump strains revenue. Upside unpreparedness costs real money; it just books the loss as missed opportunity.
- Precision theatre. Presenting a downside EBITDA of £1.037m implies a confidence no scenario possesses. Round aggressively; the decision changes at £1m versus £3m, not at the third significant figure.
- No owner for the triggers. If nobody is named to watch the indicators and call the moment, the playbook sits unread while the scenario quietly arrives. Trigger monitoring belongs in the monthly reporting cycle with a named owner.
- Treating the exercise as annual. Scenarios built once during budget season describe last autumn's uncertainties. The set must live on the same cadence as the forecast.
Building scenarios into board reporting
Scenario work changes the board conversation only if it reaches the board in usable form. What works in practice:
- Present the range, not just the base. Show the full-year outlook as base, downside and upside with the two or three driver assumptions that separate them — one slide, not an appendix of model output.
- Report the triggers as traffic lights. Each month, show where the early indicators stand against their thresholds. This converts "how confident are you in the forecast?" from a mood question into an evidence question.
- Bring the playbook, not just the pain. A downside scenario presented without its pre-agreed actions invites panic; presented with them, it demonstrates control. Boards fund prepared management teams.
- Keep scenario labels stable. If "Downside 2" means something different each quarter, the board cannot track whether the world is drifting towards it. Stable definitions, refreshed numbers.
Done this way, scenario planning becomes the connective tissue between the forecast and governance: the board sees not only what finance expects, but what finance is watching and what it will do. That is what "no surprises" actually means.
If you want this capability without a six-month modelling project, our scenario planning practice builds the driver model, the scenario set and the trigger framework with your team — and leaves you running a process the board trusts.