Rolling Forecast: What It Is, How It Works, and When to Use One

I have never understood why December 31 gets so much authority over financial planning.

Apparently the calendar changes and Finance is supposed to lose interest in what happens next.

It’s October. We have a three-month forecast remaining.

Very useful.

Management is making hiring decisions that will affect next summer, Sales is talking about deals that won’t close until February, and we’re considering an investment with an 18-month payback.

But don’t worry.

We have excellent visibility through Christmas.

This is one reason I like rolling forecasts.

A rolling forecast keeps a consistent period of financial visibility in front of the business instead of allowing the forecast horizon to get shorter as the year progresses.

Done well, it’s one of the more useful tools FP&A can give management.

Done poorly, it’s the same forecast with more columns.

I’ve seen both.

What Is a Rolling Forecast?

A rolling forecast is a financial forecast that continuously extends forward as each reporting period ends.

Instead of forecasting only through the end of the fiscal year, you maintain a consistent forecast horizon.

For example, suppose a company uses a 12-month rolling forecast.

In January, the forecast covers February through the following January.

After February closes, you add another month to the end.

Now it extends through February of the following year.

The horizon keeps moving.

You can use:

• a 12-month rolling forecast,
• an 18-month rolling forecast,
• a 24-month rolling forecast,
• or another period appropriate for the business.

Some companies forecast monthly.

Others forecast quarterly.

The exact configuration matters less to me than the reason you’re doing it.

A rolling forecast should give management enough forward visibility to make decisions.

If we’ve simply added six columns nobody looks at, I wouldn’t call that progress.

I’d call it a wider spreadsheet.

Rolling Forecast vs. Traditional Forecast

A traditional annual forecast often runs through the end of the fiscal year.

Suppose your fiscal year ends in December.

In January, management has nearly 12 months of forward visibility.

By June, it has six.

By October, three.

This has always struck me as slightly strange.

The closer we get to the end of the year, the less interested our planning process becomes in the future.

The business does not share this limitation.

A rolling forecast solves that by maintaining a constant horizon.

If management needs 12 months of visibility to make decisions, it should have approximately 12 months of visibility in January and October.

That’s the basic idea.

Rolling Forecast vs. Budget

A rolling forecast also shouldn’t replace the budget simply because it is newer.

The two can coexist.

The budget establishes the operating plan, targets and resource commitments.

The rolling forecast updates management’s expectations as actual results and new information become available.

I care about this distinction because companies have a habit of turning every financial planning tool into the same thing eventually.

Budget.

Forecast.

Target.

Stretch target.

Latest estimate.

Operating plan.

Before long there are six numbers in the meeting and someone asks which one we’re actually talking about.

A rolling forecast should answer:

Given what we know today, what do we currently expect to happen over the next 12, 18 or 24 months?

Keep that job clear.

Why Use a Rolling Forecast?

The biggest advantage is obvious:

You can see farther ahead.

But I think the more important benefit is what that does to management behavior.

Consider a company entering October.

With a calendar-year forecast, everyone is concentrating on Q4.

That makes sense.

But decisions being made today may affect:

Q1 hiring,

next year’s capacity,

renewals,

cash,

capital expenditures,

pricing,

or customer growth well beyond December.

A rolling forecast forces those effects into view earlier.

That’s useful because one of FP&A’s jobs is to make consequences visible before they become actuals.

Once they’re actuals, Accounting has them covered.

A Rolling Forecast Forces You to Stop Treating January Like a Surprise

I’ve been through enough annual planning cycles to appreciate this particular absurdity.

Everyone knows January is coming.

There is documentation.

Calendars have been available for some time.

Yet companies still arrive near year-end acting mildly surprised that they need a financial view of next year.

The annual budget process often becomes responsible for solving that problem.

Which means Finance is trying to close the current year, build next year’s budget, support the business, prepare board materials and possibly remember that holidays exist.

A rolling forecast doesn’t eliminate annual planning.

But it can mean the company isn’t starting from zero.

By the time planning begins, Finance already has a current forward-looking view.

That’s a much better conversation than opening last year’s budget and changing a few percentages.

How Far Should a Rolling Forecast Go?

There isn’t one correct answer.

This depends on the business and the decisions management needs to make.

A 12-month rolling forecast may work well for companies with relatively short planning cycles.

An 18-month forecast can provide additional visibility for hiring, investments and capacity decisions.

A 24-month forecast may make sense when major investments or longer sales cycles require more planning.

I wouldn’t choose 18 months because another company uses 18 months.

I’d ask:

How far ahead do we need to see something before we can do anything about it?

That’s the more useful question.

If it takes nine months to build capacity, a six-month forecast has an obvious problem.

If most decisions can be changed in 30 days, maintaining detailed monthly forecasts three years out may be an equally obvious waste of everyone’s Tuesday.

Forecast to the decision horizon.

How Detailed Should a Rolling Forecast Be?

This is where Finance can hurt itself.

Someone decides more visibility is good.

Then more detail must also be good.

Soon we have an 18-month forecast containing office supplies by department.

I have questions.

The further out the forecast goes, the less precise it should generally become.

Near-term periods may deserve considerable detail because we know more.

Further periods can often rely more heavily on business drivers and broader assumptions.

I think of forecast detail like eyesight.

The thing three feet in front of you should be pretty clear.

The thing 18 miles away probably shouldn’t have individual buttons on its shirt.

Pretending otherwise doesn’t improve visibility.

It improves false precision.

Step 1: Decide What the Forecast Is For

Before building anything, ask what decisions the rolling forecast needs to support.

Hiring?

Cash?

Capacity?

Revenue planning?

Investment?

Board visibility?

Capital allocation?

Cost management?

Different answers may produce different forecast designs.

I don’t start with:

How many months should the spreadsheet have?

I start with:

What does management need to see early enough to act?

That question tends to save a surprising amount of Excel.

Step 2: Choose the Right Forecast Horizon

Once you understand the decisions, choose the horizon.

Let’s say management needs approximately 12 months of forward visibility.

Maintain 12 months.

When one month closes, add another.

Simple.

The mistake is assuming the longest possible forecast is automatically the most sophisticated.

I can forecast 2047 if you’d like.

I cannot promise the number will be especially useful.

Forecast length should reflect decision usefulness, not ambition.

Step 3: Identify the Business Drivers

A rolling forecast works much better when it’s built around business drivers rather than endless line-item assumptions.

Revenue might depend on:

customers,

pricing,

volume,

retention,

pipeline,

conversion,

utilization,

or capacity.

Payroll might depend on:

headcount,

start dates,

compensation,

bonuses,

and benefits.

Gross margin may depend on:

mix,

pricing,

labor,

input costs,

or productivity.

The model should reflect how the business actually works.

This matters even more with a rolling forecast because you’ll update it repeatedly.

If every update requires manually touching 900 cells, Finance will eventually stop enjoying the process.

Possibly during the first update.

Step 4: Make Assumption Ownership Clear

I don’t want Finance inventing every assumption.

Sales should know something about pipeline.

HR should know something about hiring.

Operations should understand capacity.

Marketing should understand planned spend.

Customer Success should know where retention is getting interesting.

Finance’s job is to connect those operating assumptions to financial outcomes and challenge them where necessary.

This requires actual conversations.

I realize that’s less satisfying than buying another platform.

Unfortunately, software has yet to completely eliminate the need for coworkers.

We’re all coping.

Step 5: Establish a Forecasting Cadence

How often should you update the rolling forecast?

Again, it depends.

Monthly works well for many businesses.

Quarterly may be enough for others.

Highly volatile environments may require more frequent updates to particular areas.

I care more about having a useful cadence than an impressive one.

I’ve seen monthly forecasting processes consume so much time that Finance spends most of the month preparing to predict the next one.

That’s not maturity.

That’s cardio.

The process should be light enough to repeat and rigorous enough to trust.

Step 6: Replace Forecast With Actuals

As each period closes, replace forecast values with actual results.

Then add another period at the end.

This sounds mechanical because it is.

But don’t just replace the numbers and move on.

Look at what happened.

Where was the forecast wrong?

Why?

Which assumptions changed?

Did the business behave differently?

Did information arrive late?

Was our driver wrong?

Was the timing wrong?

This is how the forecast gets better.

Not because Finance becomes psychic.

Because it learns.

Step 7: Don’t Automatically Carry Old Assumptions Forward

Rolling forecasts can develop a particular bad habit.

Because you’re extending the model continuously, it’s easy to keep copying assumptions forward.

Revenue growth.

Hiring pace.

Margins.

Churn.

Pricing.

Suddenly an assumption that was reasonable nine months ago is still quietly running the business three quarters later.

Every so often I want somebody to stop copying cells and ask:

Do we still believe this?

I consider that one of the more productive questions available in Finance.

It’s also remarkably inexpensive.

Step 8: Use Scenarios Where Uncertainty Matters

A rolling forecast doesn’t eliminate uncertainty.

It gives you more places to put it.

I don’t need three scenarios for every line item.

I want scenarios around uncertainties that could change a management decision.

What if hiring accelerates?

What if revenue growth slows?

What if a large customer churns?

What if pricing improves?

What if the product launch moves three months?

Then show me what happens.

More importantly, tell me what management would do.

I’ve said this before because I keep seeing it:

A scenario without a decision attached is mostly an alternate future we’ve paid Finance to worry about.

I can worry for free.

I’m very efficient at it.

Step 9: Separate Near-Term Confidence From Long-Term Uncertainty

One of the things I like about rolling forecasts is that they force a useful conversation about confidence.

I might have reasonably strong confidence in next month’s payroll.

I have less confidence in revenue 14 months from now.

That’s normal.

The model should acknowledge it.

Not every period deserves the same level of precision.

This also affects how I present the forecast to management.

Near-term periods can be discussed in more detail.

Longer-term periods are often better framed around ranges, drivers and scenarios.

A forecast becomes more credible when Finance admits what it doesn’t know.

I find this is also useful outside work, although my children remain unimpressed when I describe household plans as having “a range of potential outcomes.”

Step 10: Make the Rolling Forecast Part of Management

This is the part companies sometimes miss.

They build the model.

Finance updates it.

Finance reviews it.

Finance sends it around.

Everyone has now successfully participated in forecasting except the people making the decisions.

A rolling forecast should show up in management conversations.

What changed?

Why?

Where are we heading?

Where are we off plan?

What risks are emerging?

What opportunities appeared?

What decisions need to change?

If the rolling forecast isn’t affecting hiring, spending, investment, pricing, capacity or other business decisions, I’d question why we’re maintaining it.

Forecasting isn’t valuable because the spreadsheet keeps moving.

It’s valuable because management does.

What Does a Rolling Forecast Example Look Like?

Imagine a company maintains a 12-month rolling forecast.

At the end of March, the forecast covers:

April through the following March.

March closes.

Actual March results replace the March forecast.

Finance updates the business assumptions.

A new April is added at the far end.

Now management still has approximately 12 months of forward visibility.

Suppose Sales also reports that customer pipeline has weakened.

Finance updates expected new business.

Revenue declines.

Hiring requirements change.

Cash changes.

Management can now evaluate whether planned hiring should continue at the same pace.

That’s the point.

The extra month isn’t the value.

The earlier decision is.

Rolling Forecast Advantages

When implemented well, rolling forecasts can provide several benefits.

Consistent Forward Visibility

Management doesn’t lose planning horizon as the fiscal year progresses.

Faster Response to Change

New information can be incorporated into expectations instead of waiting for the next annual budget.

Better Decision Support

Management can see the longer-term financial effects of decisions earlier.

Stronger Connection Between Operations and Finance

Regular updates force Finance and business leaders to discuss the assumptions behind the numbers.

Less Dependence on the Annual Budget

The annual plan remains useful, but it doesn’t have to carry the entire burden of forward-looking financial planning.

I particularly like that last one.

The annual budget has enough problems without asking it to predict what the company will believe 11 months later.

Rolling Forecast Disadvantages

I don’t think rolling forecasts are automatically better.

They have costs.

They Can Create More Work

If the process is poorly designed, Finance has simply volunteered to do annual planning 12 times a year.

I would decline.

They Can Create False Precision

More months do not mean more knowledge.

A detailed 18-month forecast may look impressive while relying heavily on assumptions nobody could possibly know with that precision.

Management Can Become Obsessed With Constant Reforecasting

Not every new piece of information requires a new forecast.

Sometimes businesses need time to operate.

They Can Blur Accountability

If the forecast changes constantly, people may lose sight of the original plan.

That’s why I still want the budget preserved as a separate baseline.

Rolling doesn’t mean rewriting history.

When Should a Company Use a Rolling Forecast?

I’d consider one when:

• the business changes quickly,
• revenue is volatile,
• management makes decisions with long lead times,
• hiring decisions have meaningful future impact,
• cash visibility matters,
• annual budgets become stale quickly,
• investors or boards need continuing forward visibility,
• the company is growing rapidly,
• or management frequently asks questions beyond the current fiscal year.

I would be less enthusiastic if the company doesn’t yet have a reliable basic forecasting process.

A rolling forecast won’t repair weak assumptions.

It won’t fix bad data.

It won’t make business leaders communicate with Finance.

It won’t solve unclear ownership.

Those problems will simply roll forward too.

Very efficiently.

Do You Need Rolling Forecast Software?

Maybe eventually.

Not automatically.

I wouldn’t buy software because somebody decided the company needs a rolling forecast.

First prove that the process works.

Define the drivers.

Define ownership.

Establish the cadence.

Understand the decisions.

Figure out what needs to integrate.

Then determine whether Excel, an existing planning system or a new platform makes sense.

Technology can make a good process faster.

It can also make a confused process faster.

Finance occasionally learns this after an implementation.

Usually an expensive one.

How I Know a Rolling Forecast Is Working

I don’t judge it by the number of months.

Or the sophistication of the model.

Or whether somebody says “dynamic planning” in the meeting.

I watch the conversations.

Does management learn about problems earlier?

Do assumptions get challenged?

Does Finance understand why the forecast moved?

Are operating leaders involved?

Are decisions changing because of what the forecast shows?

Does the CFO have better visibility?

Are there fewer surprises?

That’s the test.

A rolling forecast should give management more time.

Time to hire.

Time to slow hiring.

Time to adjust spending.

Time to solve a cash problem.

Time to respond to weaker demand.

Time to invest when something is working better than expected.

The value isn’t knowing the future.

Finance has not received that upgrade yet.

The value is seeing enough of it early enough to do something.

The Calendar Shouldn’t Decide How Far Ahead Finance Can See

That’s ultimately why I like rolling forecasts.

A fiscal year is useful.

Accounting needs one.

Budgets need boundaries.

Boards need reporting periods.

But management decisions don’t politely stop at year-end.

The company keeps going.

The consequences of today’s decisions keep going too.

FP&A should be able to see them.

So when someone asks me whether a company should use a rolling forecast, I don’t start with whether 12 months or 18 months is considered best practice.

I ask something else:

How far ahead does management need to see in order to change the outcome?

Build the forecast around that.

And if the answer happens to extend beyond December 31, I think the spreadsheet can handle it.