Budget, forecast or rolling forecast: which should we run?

Ask five people in a startup what the difference is between the budget and the forecast and you will get five answers, most of which amount to the same spreadsheet being called different things in different meetings.

The confusion is not academic. It produces a common and expensive failure: a plan that gets quietly revised every month so it always matches reality, which means nobody can ever say whether the company hit its targets or not. Variance analysis becomes impossible, accountability dissolves, and the annual planning exercise starts to feel pointless, because it is.

The three things are genuinely different, they answer different questions, and a company of thirty people does not need all three.

The budget: a commitment that does not move

A budget is set before the year starts and fixed for the year. It says what the company intends to spend and expects to earn, and it stays where it is.

The fixed part is the whole point. If the budget moves whenever reality does, the difference between plan and actual has no meaning, and the difference is where the information lives. A department that spent 30% more than budget has something to explain. If the budget had already been adjusted upwards twice, there is nothing to explain and nothing was learned.

In India the budget runs April to March, which matters more than founders expect. The financial year end brings statutory audit, tax filings and often a compressed collections push in March, and a budget built on a January to December calendar will misplace all of it.

What a budget is for, concretely:

• ·Setting spending limits that managers work inside without asking every time.
• ·Giving hiring approval a shape: these roles, in these quarters, at these costs.
• ·Producing variance analysis each month, which is the mechanism by which anyone learns whether assumptions were right.
• ·Giving the board something to hold the company to, which is what makes a board meeting about performance rather than about narrative.

The forecast: what you now think will happen

A forecast is different in one respect that changes everything: it is allowed to be updated. It is the company's current best estimate of how the year ends, given what is now known.

By August, the budget set in March is out of date as a prediction. Two customers churned, one large deal closed early, hiring ran a quarter late. The forecast reflects all of that. The budget does not, and should not, because it is being used for a different purpose.

Running both is the point. The budget answers what we committed to; the forecast answers where we now expect to land; the gap between them is the conversation worth having.

Changing the budget to match the forecast destroys the only comparison that teaches you anything. Keep both, and look at the gap.

Most startups reforecast quarterly. Some do it monthly for cash and quarterly for the full P&L, which is a sensible split: cash changes weekly and needs attention, while the full-year revenue view rarely moves enough in a month to justify rebuilding it.

The rolling forecast: a horizon that never shortens

A standard forecast covers the financial year. That means in April you are looking twelve months ahead and in February you are looking at one, which is exactly backwards: the closer you get to the year end, the less visibility you have of what comes next.

A rolling forecast fixes the horizon rather than the end date. A twelve-month rolling forecast always covers the next twelve months, so each month you add one at the far end and drop the one just completed. You are never looking at a two-month runway of vision in February.

That is genuinely valuable for some companies and genuinely wasted effort for others.

• ·Worth it when decisions with long lead times come up constantly: hiring that takes a quarter to land, capacity commitments, inventory, new locations.
• ·Worth it when the business is volatile enough that a March view of December is worthless by August.
• ·Worth it when cash is tight and the question of when money runs out has to be answered continuously rather than quarterly.
• ·Not worth it when the model takes two days a month to update and the answer barely moves.
• ·Not worth it when nobody uses the output, which is the usual reason they get abandoned.

Why most rolling forecasts fail

The pattern is consistent enough to predict. A company adopts a rolling forecast, runs it enthusiastically for two months, then finds it taking three days each month, then does it late, then stops.

Three causes, all avoidable.

The first is detail. A rolling forecast built at the same line-item granularity as the budget cannot be updated monthly by a small team. It needs to be a simplified model: revenue by segment, headcount by function, a handful of cost drivers. Twenty lines, not two hundred.

The second is manual data. If updating the forecast means exporting from three systems and pasting into a spreadsheet, it will not survive a busy month. The inputs have to come from the close, automatically or nearly so, which is why a reliable monthly close is a precondition rather than a nice-to-have.

The third is that nobody makes a decision with it. A forecast that goes into a folder is work with no output. If the monthly update does not change a hiring decision, a spending decision or a cash decision, the company does not need it monthly.

What most Indian startups actually need

By stage, roughly.

• ·Pre-seed and seed: an annual plan, deliberately simple, plus a 13-week cash forecast updated weekly. At this stage cash timing matters far more than a full-year P&L view, and the plan exists mostly to make assumptions explicit.
• ·Post-seed to Series A: a proper April to March budget, monthly variance review, a quarterly reforecast of the full year, and the weekly cash forecast continuing. This is enough for almost every company of this size.
• ·Series A and beyond: the same, plus a rolling twelve-month view if the business has long lead times or real volatility, and department-level budgets with owners.
• ·At any stage where cash is tight: the 13-week forecast becomes the most important document in the company, more so than either the budget or the annual forecast.

The mistake is adopting the machinery of a larger company before the basic version works. A company that cannot close its books by the tenth of the month cannot run a monthly rolling forecast, because the inputs will not exist in time.

A worked example of the gap

Illustrative figures. A company budgets FY26-27 revenue at ₹7.2 crore, costs at ₹6.6 crore, so a small surplus.

By the September reforecast, two things have happened. Revenue is tracking at ₹6.4 crore for the year, because a large customer churned in July. Costs are tracking at ₹6.1 crore, because three hires slipped a quarter.

Three numbers now sit side by side: budget ₹60 lakh surplus, forecast ₹30 lakh surplus, and the variance between them explained by one churn and one hiring delay.

That is a useful board conversation, and none of it would exist if the budget had been quietly revised in July. The discussion is not about whether the forecast is right; it is about whether the churn was a one-off or a signal, and whether the delayed hires should now happen at all given the lower revenue.

The second point is the one founders often miss. A cost underrun that offsets a revenue shortfall is not good news; it usually means the company is growing more slowly than planned in two ways at once.

Running the variance review so it is useful

Variance analysis has a poor reputation because it is often done as a ritual: a table of differences read out and nobody's behaviour changes.

What makes it useful is a threshold and an owner. Only variances above a set size get discussed, each has a person who explains it, and each explanation ends with either an action or an explicit decision that no action is needed.

It also helps to separate the two kinds of variance. A price or volume difference tells you something about the market. A timing difference, an invoice that landed in the next month, tells you nothing and should be labelled as such rather than debated. Companies that do not separate them spend most of the meeting on noise.

One more discipline: track how accurate the forecast has been. A company that keeps a simple record of forecast against actual, month after month, learns within two quarters which parts of its plan are systematically optimistic. That is usually sales timing, and knowing it by how much is more valuable than any single forecast.

Who should own each

The budget is owned by the founder or the leadership team collectively, because it is a set of commitments rather than a finance document. Finance builds it and holds the pen; the owners of each line agree to their numbers.

The forecast is owned by finance, with inputs from sales on pipeline and from each function on hiring and spending. The distinction matters: if sales owns the revenue forecast outright it will be optimistic, and if finance builds it with no sales input it will be disconnected from what is actually happening.

The 13-week cash forecast is owned by whoever can act on it, which in most startups is the founder or the finance lead, and it should be reviewed weekly with the people who control receipts and payments.

What to do this quarter

1. 01Decide which of the three you are actually running, and name it consistently in every meeting. Half the confusion is vocabulary.
2. 02If the budget has been revised during the year, stop. Freeze the original, keep the revisions as the forecast, and report both.
3. 03Set a reforecast cadence, quarterly for most companies, and put the dates in the calendar for the whole year.
4. 04Build the 13-week cash forecast if you do not have one. It is the highest-value document of the three for any company with less than eighteen months of runway.
5. 05Start recording forecast accuracy, one line per month. It costs nothing and it changes how the next plan is built.
6. 06Only then consider a rolling forecast, and only if you can name the decision it would change.

The annual operating plan guide on this site covers how to build the budget itself for an April to March year, and the 13-week cash forecast guide covers the weekly document. This piece exists because the three get confused with each other, and the confusion is what makes the whole exercise feel like paperwork instead of a way to run the company.

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.