Back to ControlFi Library

FP&A & Reporting

Rolling Forecast: How to Build One and When to Reforecast

A rolling forecast updates the outlook monthly while the budget stays fixed: budget vs forecast, when to reforecast, a SaaS example and a free template.

November 9, 2026Bogdan Stepanov10 min read

A rolling forecast is a forecast that always looks a fixed period ahead, usually 12 months, and is updated every month after the books close. Each update replaces the month that just ended with actual results and adds a new month at the end. It's driven by a few assumptions, such as new customers, churn, price and hiring, rather than hundreds of line items.

The budget doesn't roll. It stays fixed for the year as the baseline:

  • The budget says what the company planned.
  • The rolling forecast says where it's heading now.
  • The gap between them is what founders and boards need to discuss.

This guide covers:

  • the difference between a budget, a forecast and actuals
  • how a rolling forecast works month to month
  • when to do a full reforecast
  • how to build one in six steps
  • a worked example for a SaaS company after its March close

The Excel template does all of it for your own numbers.

Free rolling forecast template

This free Excel template keeps your budget fixed and rolls the forecast forward each month. You:

  1. Enter your budget once.
  2. Add actuals after each close.
  3. Update a handful of drivers for the months ahead: new customers, churn, revenue per customer, headcount and spending.

It then shows:

  • the full-year latest estimate against budget, with each variance marked favorable or unfavorable
  • the next 12 months of revenue, operating loss and cash
  • the lowest cash point and your runway
  • the first month cash would run out without new funding

A change log records what you changed and why.

It comes pre-filled with the worked example below, so you can see how it works before entering your own numbers.

Download the template (.xlsx)

Budget vs forecast vs actual

The three numbers answer different questions, and each has a job:

Budget Forecast Actual
Question it answers What did we plan? Where are we heading? What happened?
When it's set Once a year, before the year starts Updated monthly or quarterly After each month-end close
Does it change during the year? No, except after a major event such as a new funding round Yes, every update No, once the books are closed
Used for Targets, approvals, measuring performance Decisions: hiring, spending, when to raise Reporting, and the starting point for the forecast
Compared with Actuals, in budget vs actual reports The budget, to see the full-year gap The budget, and last month's forecast

The most common mistake is merging the first two. If the budget is overwritten every time the outlook changes, nobody can tell whether the company is executing well or has just lowered the bar. Keep the budget fixed and put the forecast next to it, as we explain in our guide to the annual operating plan.

How a rolling forecast works

A rolling forecast has three parts that move together each month.

  1. Closed months are actuals. After the March close, January to March show what actually happened, and the forecast for those months is discarded.
  2. The next 12 months are forecast from drivers. Revenue is customers × revenue per customer, payroll is headcount × cost per person, and so on. Change a driver and the whole forecast updates.
  3. The window moves forward. After the March close, the forecast runs from April to the following March. After the April close, it runs from May to the following April. You always see a full year ahead, including past year-end, which a static budget never shows in the second half.

Two numbers come out of every update:

  • The latest estimate for this year: actuals to date plus forecast for the remaining months, compared with the budget.
  • The next 12 months of cash: the lowest point, and when cash would run out without new funding.

A rolling forecast built this way is a driver-based version of the 3-statement model's P&L and cash, updated monthly. For week-by-week cash, add a 13-week cash flow forecast underneath it.

Worked example: reforecasting after the March close

Take the $4 million ARR SaaS company from our budget vs actual guide. It started the year with $6 million in the bank. The March close showed three things the budget hadn't assumed:

  • churn of 2 customers a month, against about 1 in the plan
  • revenue per customer of $1,472, against $1,500, after renewal discounts
  • two hires starting a month late

The founder updates the drivers for the rest of the year:

  • Customers: 5 new customers a month, plus the 2 slipped deals in April, and churn of 2 a month.
  • Price: revenue per customer of $1,480 for the rest of 2026, with a list-price increase to $1,520 from January 2027.
  • Hiring: slowed to reach 38 people by December instead of 40.
  • One-offs: $45,000 of Series B legal fees spread over April and May.

Full year 2026: budget vs latest estimate

Line Budget Latest estimate Variance F/U
Total revenue $4,513,500 $4,416,291 −$97,209 (−2.2%) U
Gross profit $3,520,530 $3,439,734 −$80,796 (−2.3%) U
Total operating expenses $6,857,000 $6,735,300 −$121,700 (−1.8%) F
Operating loss −$3,336,470 −$3,295,566 +$40,904 (+1.2%) F
Paying customers, December 257 254 −3 U
ARR, December $4,626,000 $4,511,040 −$114,960 (−2.5%) U
Headcount, December 40 38 −2
Cash, December 31 $2,663,530 $2,704,434 +$40,904 F

On profit, the year now looks better than budget: slower hiring saves more than the lower revenue costs. A static budget vs actual report would show a company under control.

The forecast shows the real issue. December ARR is $114,960 behind plan, and ARR growth is what a Series B investor will look at. That's the conversation the board needs to have now, in April, not in December.

The next 12 months: April 2026 to March 2027

Measure Value
Cash at March 31 $5,256,851
Average monthly net burn $285,203
Runway at the March close (cash ÷ average burn over the next 3 months) 19.3 months
Lowest cash in the window $1,834,416, in March 2027
First month cash would go below zero, with no new funding October 2027

Nineteen months of runway sounds comfortable. But without the Series B, cash runs out in October 2027. Most founders want to close a round with at least five or six months of cash left, which here means by about April 2027. Allowing four to six months for the process, the raise needs to start in the fourth quarter of 2026, with ARR currently behind plan.

That's a decision a rolling forecast surfaces in April, eight months before the budget would have shown a problem. The burn multiple is the second check: at roughly $3.4 million of operating loss over the next 12 months, against about $0.9 million of ARR added, it's running at nearly 4x. That's high for a Series A company, so investors will ask about it.

When to reforecast

A rolling forecast is updated every month, but most months only a few drivers change. A full reforecast, revisiting every assumption and re-presenting the year to the board, is worth doing when something material changes:

  1. Revenue drifts from plan for two months running. For example, new ARR more than 20% below plan for two consecutive months, the same trigger we use in the annual operating plan guide.
  2. Funding changes. A round closes, slips or comes in smaller, or a debt facility is added.
  3. A large customer is won or lost. For a startup, a single customer can move the year.
  4. The hiring plan changes. Accelerating or freezing hiring changes burn for the next 12 months.
  5. Pricing or packaging changes. That moves revenue per customer for every future month.
  6. Every quarter, as a minimum. Even with no single trigger, small drifts add up.

The budget stays fixed through all of these. Reset it only after a change so large that the original plan no longer makes sense as a baseline, such as a new funding round that changes the plan itself.

How to build a rolling forecast in six steps

  1. Pick the horizon and the rhythm. Twelve months ahead, updated every month within a few days of the close. Use 18 months if you're within a year of a fundraise, so you can see past the round.

  2. Forecast from drivers, not line items. For a SaaS company, that means:

    • new customers and churn
    • revenue per customer
    • headcount and cost per person
    • a few large spending lines

    Small lines can grow at a set rate.

  3. Lock the actuals. Closed months come straight from the books and are never forecast again.

  4. Update only what changed, and write down why. A short change log, like "churn reset to 2 a month after Q1 data", is what makes the forecast credible to a board.

  5. Roll the window. Add one month at the end each time you close one at the start.

  6. Compare and decide. Look at the full-year latest estimate against budget, then the next 12 months of cash: the lowest point, the runway and when the next raise has to start. End each update with the decisions it changes.

Five rolling forecast mistakes

  1. Overwriting the budget. The forecast moves and the budget doesn't. Without the fixed baseline, every variance disappears.
  2. Forecasting the target. A forecast is the honest best estimate, not the number the board wants. If the forecast always matches the plan, nobody trusts either.
  3. Too many line items. A forecast with 200 lines takes days to update and gets skipped. Ten or fifteen drivers usually explain most of the result.
  4. Changing everything every month. Small movements in drivers are noise. Change a driver when the evidence changes, and record why.
  5. Forecasting profit and ignoring cash. A forecast that shows a better-than-budget year can still hide a cash-out date. Always end with cash, runway and the date the next raise must start.

Frequently asked questions

Clear answers to the questions founders ask most often.

A forecast that always covers a fixed period ahead, usually 12 months, and is updated every month. Each update replaces the month just closed with actuals and adds a new month at the end, so the window keeps moving forward.

A budget is the plan, set once a year and kept fixed as the baseline. A forecast is the current best estimate of where the company will land, updated as new information comes in. Compare actuals with the budget to judge performance, and use the forecast to make decisions.

A full update of the forecast for the rest of the year, done when something material changes, such as a revenue shortfall, a funding event or a hiring change. In a rolling forecast, small updates happen monthly and a full reforecast happens when a trigger is hit, or at least quarterly.

Monthly, within a few days of closing the books. Most months only a few drivers change. Revisit every assumption at least once a quarter.

Twelve months is standard for startups. Use 18 months if you're within a year of raising, so the forecast shows the period after the round. For week-by-week cash, use a 13-week cash flow forecast alongside it.

No. Keep both. The budget is the fixed baseline that shows whether the company is executing well. The forecast shows where it's heading. Replacing the budget with the forecast removes the comparison.

The full-year number made from actuals for the months already closed plus the forecast for the rest of the year. It's compared with the budget to show how the year is likely to end.

Yes, but for cash, week by week. It rolls forward one week at a time and catches short-term cash problems. A 12-month rolling forecast covers revenue, costs and cash month by month, and shows longer-term runway and fundraising timing.

The short version

  • A rolling forecast always looks 12 months ahead and is updated after every monthly close.
  • Keep the budget fixed and compare the forecast's latest estimate with it.
  • Forecast from a few drivers: customers, churn, price, headcount and large spending lines.
  • Do a full reforecast when something material changes, and at least every quarter.
  • End every update with cash: the lowest point, the runway and when the next raise has to start.
  • A better-than-budget profit can hide a growth or cash problem. The forecast shows it months earlier.

If you want a rolling forecast built and updated every month alongside your board pack, our fractional CFO services page explains how ControlFi works.

Sources:
The worked example, triggers and template are illustrative and prepared by ControlFi.

About the Author

Bogdan Stepanov

Founder, ControlFi

Bogdan has spent over nine years in M&A, investing and private markets, advising on 30+ cross-border transactions and private financings with $700M+ in aggregate value across EMEA, North America and APAC. He has built and reviewed 100+ financial models.

Connect on LinkedIn