How do you budget when revenue is unpredictable?

The annual planning advice all assumes the same thing: that you can produce a revenue forecast good enough to plan costs against. For a subscription business with a year of history and low churn, that is reasonable.

For a great many businesses it is not. A services firm whose year depends on four deals closing. A company selling to government or large corporates where timing is outside your control. A usage-based product where two accounts decide the month. A business in a market that did not exist eighteen months ago.

Founders in that position tend to do one of two things, and both are bad. They produce a confident-looking plan everyone knows is fiction, and then ignore it. Or they do not plan at all, and spend reactively, which feels prudent and is how a cost base grows without anyone approving it.

There is a third approach, and it starts by giving up on the thing that cannot be done.

Plan the spending, not the revenue

The insight is that a budget's real job is not to predict revenue. It is to decide what you will spend, and a revenue forecast is only the input that usually informs it.

So invert it. Ask what level of spending is defensible against revenue you are genuinely confident about, commit only that, and hold the rest of the plan as decisions that get taken later when more is known.

That produces a plan with a different shape: a committed base that is safe in almost any outcome, and a set of conditional additions each attached to something observable.

A plan built on revenue you cannot forecast is a wish. A plan built on spending you can control is a decision.

Three tiers of revenue confidence

The first piece of work is separating your revenue into what you actually know, and it is usually more informative than the forecast it replaces.

• ·Committed: contracted, signed, with a minimum. Subscriptions in force, retainers, annual commitments, the contracted portion of multi-year deals. This is revenue that arrives unless something goes wrong, and it is the only tier the cost base should be built on.
• ·Probable: repeat business from customers who have bought consistently, renewals that have not been signed yet, and pipeline weighted by how often deals at that stage have actually closed. Real, and not certain.
• ·Possible: new logos not yet in the pipeline, expansion nobody has discussed, the large deal everyone is hoping for. Worth pursuing, and worth nothing in a plan.

Most companies doing this for the first time are surprised by how small the committed tier is, and that surprise is the point. A business planning a cost base against a number that turns out to be 40% committed and 60% hoped for has learned something important before writing a single budget line.

Build the cost base in the same three tiers

1. 01The floor. What the business costs to exist: rent, core team, infrastructure, compliance, insurance. This has to be covered by committed revenue plus whatever cash you are willing to spend, and if it is not, that is the finding and nothing else in the plan matters yet.
2. 02The committed plan. Hires, tools and spending you will do regardless, funded by the committed tier. Approve these now and get on with them.
3. 03The conditional plan. Everything else: the growth hires, the marketing step-up, the second location. Fully specified, costed, ready to execute, and not approved. Each one attached to a trigger.

The third tier is what makes this work. It is not a list of things you might do. It is a set of decisions that have already been made and are waiting for a condition, which means they can be executed the week the condition is met rather than debated for a month first.

Triggers, not dates

A trigger is a specific, observable condition that releases a piece of conditional spending. Written in advance, agreed by whoever would have to approve it, and checked monthly.

• ·If committed revenue for the next six months exceeds ₹X, the two sales hires are approved.
• ·If the Q2 renewal cohort signs at or above 85% by value, the marketing budget increases by ₹Y a month.
• ·If we close two of the four enterprise deals, the second location proceeds.
• ·If cash falls below ₹Z, the conditional plan is suspended in full and we revert to the floor.

Two properties matter. A trigger has to be checkable by someone who was not in the room, so no condition that depends on how anyone feels about the pipeline. And every trigger that releases spending needs a matching one that withdraws it, or the plan only ratchets upwards.

A worked example

Illustrative figures for a fictional services firm. Last year's revenue was ₹9.4 crore. Of next year's, ₹4.8 crore is contracted retainers and signed work, ₹2.6 crore is weighted pipeline, and management hopes for another ₹3 crore of new business.

• ·The floor is ₹4.1 crore a year: core team, premises, infrastructure and compliance. Covered comfortably by committed revenue.
• ·The committed plan takes total spending to ₹5.4 crore, still inside the ₹4.8 crore of committed revenue plus a modest draw on cash. This is approved now.
• ·The conditional plan is a further ₹2.2 crore: four hires, a marketing step-up and an office move. Each is costed and each has a trigger.
• ·One trigger: the two delivery hires release when committed revenue for the following two quarters passes ₹3.2 crore. Another: the office move proceeds only if headcount actually passes 40, rather than when the plan says it will.

Compare that with the conventional version, which would have budgeted against ₹10.4 crore of expected revenue, committed the full ₹7.6 crore of cost in April, and spent the year discovering whether the revenue arrived.

The tiered version is worse in one specific way, and it is worth naming: if the revenue does arrive, the company is slower to spend against it, and some of the growth those hires would have produced does not happen. That is the cost of the approach, and it is the right trade for a business that cannot forecast, because the alternative failure mode is considerably more expensive.

Getting the committed number right

Everything here rests on the committed tier, so it is worth being strict about what goes in it. The temptation to include things that are nearly certain is exactly what the approach exists to resist.

• ·A contract with a notice period is committed only for the notice period, not for its full term. A twelve-month agreement cancellable on thirty days is one month of committed revenue.
• ·A customer who has always renewed is probable, not committed, until they have signed. Long relationships end, and they usually end at renewal.
• ·A verbal commitment from someone senior is probable. A purchase order is committed.
• ·Usage above a contracted minimum is probable however reliably it has recurred. Only the minimum is committed.
• ·And a customer in financial difficulty is not committed regardless of what they have signed, which is a judgement worth making explicitly rather than ignoring.

Applied strictly, the committed tier in most businesses is smaller than the first attempt at it. That is the number worth knowing, and a company that plans its floor against an inflated version of it has reintroduced the problem this whole approach was meant to remove.

What to report monthly

The reporting changes shape too, and it becomes considerably more useful.

• ·Committed revenue for the next six and twelve months, which is the number this whole approach turns on. Watch it monthly, because it moves.
• ·The floor, the committed plan and the conditional plan as three separate cost lines, so everyone can see which is running.
• ·Which triggers have been met, which are close, and which have moved away.
• ·Cash against the point at which the conditional plan suspends.
• ·And actual revenue against each of the three tiers, so you learn over time how well you judge probable.

The last one matters more than it looks. After four quarters you will know what share of your probable tier actually lands, and that ratio is more useful than any forecast, because it can be applied to next year's probable tier with some confidence.

Telling the team about a conditional plan

This approach has a people problem that a conventional budget does not, and ignoring it is how it fails in practice.

A team told that four hires are in the plan will behave as though four hires are coming. When two of them are conditional and the condition is not met, that reads as a cut rather than as the plan working correctly, and it damages trust in exactly the way an actual cut would.

So the conditional part has to be communicated as conditional from the start, with the trigger named. We are hiring two people now. Two more are approved and will start when committed revenue for the next two quarters passes this figure, which we are currently below and expect to reach in the second quarter.

That is a more demanding thing to say than a simple headcount plan, and it has a useful side effect: it tells the whole company what number actually matters, which is frequently the first time anyone outside the leadership team has been told.

The version that does not work is holding the conditional plan privately and announcing each release as though it were a new decision. People notice the pattern, and it reads as a company that cannot plan rather than one planning carefully.

What this does not solve

• ·It does not make the revenue more predictable. Nothing on this page does. It stops the spending from depending on the prediction.
• ·It does not work if the floor exceeds committed revenue by a wide margin. That is a company relying on hoped-for revenue to cover its existence, and the answer is to reduce the floor or accept that cash is funding the gap deliberately.
• ·It is harder to explain to investors, who are used to a single-number plan. The answer is to show the committed base, the triggers, and what the year looks like at two or three levels of outcome. Most find it more credible than a confident forecast, because it demonstrates that the downside was considered.
• ·And it needs discipline that a conventional budget does not. A conditional plan that gets released because everyone feels optimistic is just a budget with extra steps.

The underlying shift is small and it is genuinely difficult for most founders: accepting that you do not know what next year's revenue will be, writing that down, and planning anyway. Most of the damage done by planning in unpredictable businesses comes from refusing to admit the uncertainty and then committing costs as though it were not there.

The scenario planning piece on this site covers the same logic applied to specific uncertainties rather than to the whole plan, and the annual operating plan guide covers the conventional version for businesses that can forecast. This one is for the businesses that cannot, which is more of them than the planning literature admits.

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.