How do you plan for scenarios without pretending to predict?

Almost every startup plan has a best case, a base case and a worst case. In most of them, the three cases are the same model with the growth rate set to 8%, 5% and 3%, and nobody has ever looked at the first or the third again after the board meeting.

That is not scenario planning. It is sensitivity analysis with a more impressive label, and it has a specific failure mode: it gives a false sense that uncertainty has been addressed, which makes it slightly worse than having no cases at all.

Real scenario planning answers a different question. Not how much might the number vary, but what would we do differently if the world turned out to work another way.

What a scenario actually is

A scenario is a coherent set of conditions that could plausibly hold together, with the consequences followed through.

The difference from a sensitivity is that a scenario changes several things at once, because in reality they move together. If enterprise sales cycles lengthen, that does not only reduce revenue. It also means cash comes later, the sales team needs more months to hit quota, the marketing spend per closed deal rises, and the hiring plan built on that revenue is no longer affordable.

A sensitivity would have moved one growth number. A scenario moves all of them, coherently, and ends up somewhere the single-variable version never reaches.

If your three cases differ only by a growth rate, you have not built scenarios. You have built one plan and moved a slider.

Start from the assumptions that carry the most weight

The practical starting point is a short list: which assumptions, if wrong, would most change what the company does?

For most early-stage companies the candidates are fewer than founders expect.

• ·How long it takes to close a customer, and whether that is stable as you move up market.
• ·Whether the next round happens, and when.
• ·Retention, particularly whether recent cohorts hold at the level older ones did.
• ·What a new salesperson or a new location produces, and how long the ramp takes.
• ·For businesses with inventory or capacity, whether demand arrives before the commitment has to be made.
• ·For regulated or platform-dependent businesses, whether a rule or a commission rate changes.

Rank them by two things: how uncertain each is, and how much the plan changes if it moves. The ones that are both uncertain and consequential are your scenarios. Usually there are two, occasionally three. More than three and nobody will use them.

Build each scenario as a world, not a column

Take the assumption, imagine it is true, and follow it through every part of the business. The discipline is to keep asking what else that would mean.

An example of the chain, for a company whose key uncertainty is fundraising:

1. 01The round does not close within nine months.
2. 02So the hiring plan stops at current headcount, and the two senior roles planned for Q3 do not happen.
3. 03So the sales capacity assumed in the revenue plan is not there, and revenue growth slows by roughly the contribution of those roles.
4. 04So collections and pricing become the only levers available, and both need to start now rather than at month nine.
5. 05So the discretionary spending approved for the year is reviewed in month two, not month eight.
6. 06And the milestone the round was meant to fund has to be reachable on existing capacity, or it has to change.

By the end of that chain you have something a single number could never produce: a specific set of actions, most of which are better started early.

A worked example

Illustrative figures. A company with ₹3.2 crore of cash, ₹52 lakh of monthly revenue growing 5%, burn of ₹26 lakh, so about twelve months of runway. Their plan assumes a Series A closing in month eight.

Two scenarios matter to them.

• ·Scenario one, the round is delayed six months. Hiring freezes at current levels, growth slows from 5% to about 3% because two planned sales hires do not arrive, and burn holds at ₹26 lakh until they act. Cash runs out in month twelve with the round still four months away. The action this scenario forces is a decision point in month four, not month ten.
• ·Scenario two, enterprise deals take eight months rather than five. Revenue growth halves for two quarters while the pipeline rebuilds, cash collection moves out by about a quarter on the affected deals, and the company needs roughly ₹40 lakh more of working capital to cover the gap. The action is a change to how the pipeline is weighted, and a decision about whether to keep pursuing that segment at this stage.

Neither scenario is a prediction. Both produce a decision the company can take now: start the collections work immediately, set a month-four review with a defined trigger, and reweight the pipeline model so the forecast stops assuming a five-month cycle it has not achieved.

The part that makes it work: triggers

A scenario nobody revisits is a document. What turns it into a management tool is a trigger: a specific, observable condition that says this scenario is now the one you are in.

Good triggers are unambiguous and checkable monthly.

• ·If we have not received a term sheet by 31 January, we move to the delayed-round plan.
• ·If two consecutive months come in below ₹45 lakh, we pause the hiring plan.
• ·If the average enterprise cycle exceeds seven months across the last six closed deals, we reweight the forecast and revisit the segment.
• ·If cash falls below ₹1.8 crore, the discretionary spending list is cut in full rather than reviewed.

The value of writing these down in advance is entirely psychological, and it is large. Decisions made in a calm month are better than the same decisions made in an anxious one, and a trigger removes the argument about whether things are bad enough to act, because the condition was agreed when nobody was under pressure.

Where this goes wrong

• ·Too many scenarios. Five cases means nobody remembers any of them, and the exercise becomes a modelling project rather than a decision aid.
• ·Scenarios that are not actually different. A best case at 7% growth and a base at 5% lead to the same actions, so they are one scenario.
• ·A worst case that is mild. Most worst cases are a modest slowdown. The useful version is the one that is genuinely uncomfortable, because that is where the pre-made decision has value.
• ·No owner and no review. A scenario set built once for a board meeting and never reopened has done nothing.
• ·Confusing the scenario with the plan. The company runs one plan; scenarios tell it when to change plan.
• ·Presenting all three cases to investors as equally likely, which reads as an unwillingness to commit to a view.

The scenario nobody builds

Scenario sets are almost always about things going worse. The upside case gets a column and no thought, which is a genuine loss, because a company caught unprepared by success handles it badly too.

The questions are real ones. If demand doubles in a quarter, can you deliver it, and what breaks first? If the round is oversubscribed, what would you spend the extra money on, and would you actually want to at that price? If a large customer wants five times the volume, do you have the working capital to fund the delivery before they pay you?

That last one is the common failure. Growth consumes cash before it produces it, and a company that wins a much larger contract than planned can find itself tighter on cash three months later than it was before. A quick upside scenario, with the working capital consequences followed through, is usually the one that saves an awkward quarter.

It need not be elaborate. One paragraph, the cash consequence, and a trigger: if a single order above a certain size arrives, we work out the funding before we sign it.

What to show investors

Investors want the base case, with the assumptions behind it visible and defensible. That is the plan you are committing to, and presenting three cases with equal weight suggests you do not have a view on which is right.

The scenarios are still worth mentioning, briefly, in a specific way: here are the two assumptions this plan is most sensitive to, here is what we would do if either turned out differently, and here is the trigger we would act on. That demonstrates exactly what a good investor is assessing, which is whether the founder has thought past the happy path.

In a board pack, the fuller version belongs in an appendix, revisited when a trigger is approaching rather than every month.

Scenarios and the board

There is a version of this exercise that goes wrong in the boardroom rather than in the spreadsheet. A founder presents two scenarios, an investor seizes on the downside one, and the rest of the meeting is spent defending a case the founder themselves considers unlikely.

The way to avoid that is to be explicit about weight and about purpose. This is the plan we are running. These are the two assumptions it depends on most. Here is what we would do if either moved, and here is the trigger. We are not forecasting the downside, we are saying what we would do about it.

Framed that way, the scenarios do the opposite of undermining the plan. They demonstrate that the plan has been stress tested, which is the thing most early-stage plans conspicuously have not been.

It helps to bring the triggers back in following meetings, briefly, even when none has been hit. A board that has watched a founder check the same conditions for three quarters treats the eventual decision to act as considered rather than reactive.

How much work this actually is

Far less than the language around it suggests. The first pass for a company of thirty people is a half day, and most of that is the conversation rather than the spreadsheet.

List the assumptions. Pick the two that matter most. Write a paragraph describing each scenario as a world. Model the cash and revenue consequences roughly, in the plan you already have, using a single switch rather than three separate files. Write the triggers. Agree who checks them and when.

The modelling discipline worth adopting is to keep the assumptions in one place in the model, so a scenario is a change to a small set of inputs rather than a copy of the whole file. Three copies of a model diverge within a month, and by the second quarter nobody knows which one is current.

Reviewing it is a fifteen-minute item in the monthly meeting: have any triggers been hit, and has anything changed about which assumptions are the uncertain ones? Twice a year, usually at the half year and at planning time, rebuild the list properly, because the uncertainties that mattered in April are rarely the ones that matter in October.

That is the whole method. It is not forecasting, it does not require better prediction than anyone else has, and it produces something a single-case plan never does: a company that has already decided what it will do when the plan stops being true.

About Simplify

Simplify is a finance clarity and investment readiness practice working with founders across India, built on six years inside startups. We write about the questions founders bring before a decision, not after it.