A rolling forecast is a financial forecast that always looks the same distance ahead, usually 12 months. Each month, the company replaces the month that just closed with actual results and adds a new month at the far end of the horizon. The plan never expires, unlike a static annual budget that runs out of road every December.
That is the whole mechanism. The complication is that nearly everything written about rolling forecasts assumes an FP&A department, a planning platform, and a quarterly workshop calendar. A 10-person startup has none of those things. The founder is the FP&A team, and the forecast has to survive on 90 minutes a month.
This guide covers both halves: what a rolling forecast is and how it compares to a static budget, then a version of the cadence sized for an early-stage company. It is one spoke of our full guide to startup financial modeling, forecasting, and planning, which covers the model that sits underneath any forecast worth updating.
Rolling forecast vs static budget vs reforecast
Three planning tools get confused with each other, and the difference is mostly cadence and horizon. A static budget is set once and compared against reality until the fiscal year ends. A reforecast patches that budget partway through the year. A rolling forecast replaces the annual rhythm entirely.
| Static annual budget | Reforecast (3+9, 6+6) | Rolling forecast | |
|---|---|---|---|
| Cadence | Built once a year | Once or twice a year | Updated monthly or quarterly |
| Horizon | Fixed fiscal year, shrinks every month | Remainder of the fiscal year | Constant 12 to 18 months ahead |
| Effort | Heavy: a multi-week planning season | Moderate: a few days per cycle | Light per cycle once the model exists |
| Fit by stage | Pre-seed and early seed | Seed startups with an active board | Post-revenue seed through Series A and beyond |
The reforecast shorthand is worth decoding because it shows up constantly in board conversations. A 3+9 forecast means three months of actuals plus nine months of updated projections for the same fiscal year. A 6+6 is the mid-year version. Both are improvements on a stale budget model, but the horizon still collapses to zero in the final quarter, which is exactly when next year's hiring and fundraising decisions need a forward view.
A rolling forecast fixes that by decoupling the forecast from the fiscal year. In month one you can see 12 months out. In month eleven you can still see 12 months out. For a startup timing a raise, that constant horizon is the entire point, because runway questions never arrive conveniently aligned to a fiscal calendar.
Why most rolling forecast advice does not fit startups
Search the term and the top results come from NetSuite, Workday, Planful, and Pigment. Those are good articles for their audience, which is corporate finance teams evaluating planning software. They describe driver trees maintained by analysts, cross-functional input cycles, and rollouts measured in quarters.
None of that maps to a startup where one founder owns the spreadsheet. The concept transfers; the process does not. An early-stage company does not need departmental submissions or a consolidation layer. It needs five drivers, one sheet, and a repeatable monthly hour. The rest of this guide describes that version, and our broader SaaS founder's guide to financial forecasting covers how the forecast connects to the rest of the finance stack.
The other thing enterprise content skips is sequencing. Big companies debate static versus rolling as a transformation project. Founders just need to know when the switch is worth the effort, because adopting a rolling forecast too early wastes time that should go into finding customers.
When to graduate from a static budget to a rolling forecast
Pre-revenue, a rolling forecast is overkill. A simple annual budget plus a monthly review against actuals gives you everything a forecast would, because expenses are the only moving part and they move slowly. If that describes your company, start with a startup budget template and revisit this guide after launch.
The switch earns its keep when three signals stack up. First, you have revenue that changes month to month, so the top line is no longer a flat assumption. Second, your budget vs actuals review keeps surfacing variances above 15 or 20 percent, meaning the annual plan has stopped describing the business. Third, decisions with long lead times are landing mid-year: a hire that takes three months to close, a fundraise that takes six.
For most SaaS startups those conditions arrive somewhere between first revenue and Series A. The pattern shows up in the numbers: a plan written in January is materially wrong by May, not because the planning was sloppy, but because a startup learns faster than an annual document can absorb. At that point, stop patching the budget and let the forecast roll.
A 12-month rolling forecast example for a seed-stage SaaS
Here is the cadence in practice. Every number below is illustrative, built for a fictional 10-person seed-stage SaaS startup so the mechanics are concrete. It is not a benchmark.
The setup: the company raised a $3.2M seed, holds $2.2M in cash, and enters March 2026 with $45,000 in MRR. Monthly operating spend is $185,000, so net burn runs near $140,000. The forecast horizon covers March 2026 through February 2027, with five drivers: new MRR added, churn, hiring start dates, marketing spend, and infrastructure cost per customer.
In early April the March books close, and the roll happens. March actuals replace the March forecast column. New MRR came in at $2,100 against a forecast of $3,000, and a senior engineer accepted a month later than planned. Both facts flow into the assumptions, and March 2027 is appended to the end so the horizon stays at 12 months.
| Roll step (April update) | Forecast said | Actuals said | What carries forward |
|---|---|---|---|
| New MRR added in March | $3,000 | $2,100 | Growth assumption trimmed from 6% to 4.5% monthly |
| Gross churn | 2.0% | 1.8% | Churn assumption held at 2% |
| Engineering hire | Starts April 1 | Starts May 1 | Payroll shifts one month, burn drops $14K in April |
| Horizon | Through Feb 2027 | March 2027 added; 12-month view intact |
The payoff is what the roll reveals. Slower growth plus a delayed hire nets out to a runway of roughly 15 months instead of the 16 in the original plan, and the founder learns that in April rather than at a year-end reforecast. Recomputing that number is one division: cash on hand over average forecast net burn, and our startup runway calculator does the sensitivity math on how it moves as burn changes.
Note what the example does not do. It does not reforecast all 40 budget lines, rebuild the model, or convene anyone. One month of truth in, one month of horizon out, three assumptions touched.
A rolling forecast also does not replace the annual budget; the two coexist with different jobs. The budget stays frozen as the commitment the board approved, and the rolling forecast carries the current best estimate of what will actually happen. When the gap between them widens for three consecutive months, that is the signal to have a deliberate conversation about the plan rather than quietly drifting away from it.
The monthly cadence in 90 minutes
The roll only stays cheap if the process is fixed. The version below assumes the books close within a week or so of month end, which is the real prerequisite for rolling forecasts at any company size. Stale actuals make every downstream number fiction.
The sequence runs in five steps. Pull the closed month's actuals into the sheet. Scan the variances and write one sentence on each gap above 10 percent, which is variance analysis at founder scale. Update only the drivers the variances implicate, and resist rewriting assumptions that reality has not challenged. Append the new month at the end of the horizon. Finish by rechecking runway and one downside case, a lightweight form of scenario planning that answers the only question that matters: what happens to the raise date if growth stays at this month's rate.
Two mistakes break the cadence. The first is adding drivers until the update takes a full day, at which point it stops happening; five to eight drivers is the working ceiling for one owner. The second is treating the forecast as a report instead of a decision tool. If a roll never changes a hiring date or a spend plan, the numbers are being recorded, not used, and the same trap that makes run rate a poor crystal ball applies to any forecast nobody acts on.
Where Futureproof fits
The 90-minute version still depends on the founder doing the roll every month, and on books that close fast enough to feed it. That is the part Futureproof automates. Vic keeps the bookkeeping current, and Margo maintains the rolling forecast directly from live actuals, so the roll happens continuously instead of in a monthly sprint, with burn rate and runway recomputed as transactions land.
Founders get the constant 12-month horizon without owning the spreadsheet ritual. The full AI finance team, six agents covering bookkeeping through investor reporting, runs $1,000 per month flat. Start with Futureproof and the first roll happens without you.
FAQ
What is a rolling forecast in simple terms?
A rolling forecast is a plan that always projects a fixed window into the future, usually 12 months. When a month ends, its actual results replace the projection and a new month is added at the far end. The horizon never shrinks the way an annual budget's does.
How is a rolling forecast different from a budget?
A budget is a fixed commitment for a fiscal year, set once and measured against. A rolling forecast is a living estimate that updates every month or quarter as results arrive. Most companies keep both: the budget as the accountability baseline, the rolling forecast as the current best view of the future.
What do 3+9 and 6+6 forecasts mean?
They describe mid-year reforecasts of an annual plan. A 3+9 combines three months of actuals with nine months of revised projections; a 6+6 does the same at the half-year mark. They refresh the numbers but keep the fiscal-year horizon, so visibility still collapses as year-end approaches.
How far ahead should a rolling forecast look?
Twelve months is the standard for startups, and it matches the horizon investors ask about. Enterprises sometimes roll 18 or 24 months, but assumptions beyond a year are guesswork for an early-stage company. A constant 12-month view is enough to time hires and start fundraises early.
How often should a startup update a rolling forecast?
Monthly, immediately after the books close. Quarterly updates suit slow-moving businesses, but a startup's growth rate and burn can shift meaningfully inside a quarter. If a monthly update takes more than about 90 minutes, the model has too many drivers, not too little time.
