Rolling Forecast vs. Annual Budget: Do You Actually Need Both?

Every annual budget eventually becomes a historical document.

Sometimes it takes twelve months.

Sometimes it takes until February.

A customer delays, hiring moves, pricing changes, a launch slips, and the beautiful approved plan begins acquiring the practical usefulness of a Blockbuster membership card.

This is usually when someone says, “We need a rolling forecast.”

I agree.

Then I watch companies make the rolling forecast behave exactly like the budget, which gives Finance two processes and one problem.

I use the budget and forecast for different conversations

The budget is where I expect commitment.

What are we trying to accomplish? Where are we putting resources? What spending has been approved? What targets did management sign up for?

The rolling forecast is where I expect honesty about the current trajectory.

Given what we know now, where do we actually think revenue, margin, cash and headcount will land?

Those numbers are allowed to disagree.

In fact, the disagreement is often the useful part.

The forecast gets weird when changing it feels like surrender

I have been in enough forecast discussions to recognize the moment.

Revenue comes down and the room stops discussing the business and starts negotiating with the number.

Are we sure?

Can we leave it where it was?

What message does this send?

That teaches everyone involved that a forecast is not really an estimate. It is a target wearing casual clothes.

If rain is now likely, I want the weather forecast to say rain. The picnic having been approved in November does not change the clouds.

I am not interested in killing the annual budget

Declaring the budget dead has been fashionable for a long time.

I still find budgets useful.

They force choices. They allocate resources. They establish spending guardrails. They coordinate departments. They support compensation and board expectations.

What I do not ask them to do is remain the best estimate of the future after the facts change.

That is where the rolling forecast earns its keep.

The rolling part is less important than the changing assumptions

Technically, a rolling forecast keeps extending the planning horizon. Finish a month, add another month.

Useful.

But the real value is that the assumptions move too.

Pipeline changed. Hiring changed. Churn changed. Collections changed. Costs changed.

If the forecast takes six weeks to absorb that information, I am not especially impressed that somebody remembered to add another column on the right.

The forecast has to change early enough to affect a decision.

I usually want both, but not two giant models

For many companies, I like the annual budget as the commitment view and the rolling forecast as the current-expectation view.

I do not like maintaining two independent modeling universes unless there is a very good reason.

Actuals, business drivers and core logic should be shared where practical. The difference should live in purpose and assumptions, not in a ceremonial second spreadsheet with its own ecosystem of broken links.

Finance has a habit of solving conceptual differences with additional tabs.

Tabs are not governance.

The bridge is where management gets value

I would keep the original budget visible and bridge it to the current forecast.

What changed?

Is it timing or structural?

When did we know?

What can management still influence?

What needs a decision now?

That conversation is usually more useful than arguing over which number deserves to be called “the plan.”

If the CEO asks me today where I think the year will land, I want to answer with the latest evidence.

If the CEO asks what we originally committed to, I want that too.

Two questions. Two useful views.

They do not need to audition for the same role.

For the mechanics behind keeping the outlook current, I use the process described in my FP&A forecasting guide. The broader forecasting and planning work is where I help teams design the cadence around the business rather than whatever calendar they inherited.