Finance and FP&A · Update the outlook

Rolling Forecasts: How to Build and Update a Business Forecast

A rolling forecast keeps a forward-looking planning window in view as time passes. Each update incorporates new actual results, revises the remaining assumptions, and adds a new period at the far end. It helps managers prepare for demand, capacity, cost, and cash decisions without waiting for the next annual budget cycle.

← Finance hub · Variance Analysis · Financial Ratios

Budget, forecast, and rolling forecast: the differences

A business can use all three views. Forecasting should be an honest expectation, not an attempt to make the latest numbers equal the budget. Keep the budget visible for accountability while using the forecast for current decisions.

Choose the horizon and update cadence

Match the horizon to the decisions. A 12- or 18-month monthly outlook may support staffing and operating planning; a longer view may be needed for capital projects. Weekly cash forecasts serve a different, shorter-term decision cycle. There is no universally best window.

Example: at September 30, 2026, a next-12-month view covers October 2026 through September 2027. After October closes, the next-12-month view covers November 2026 through October 2027. October 2026 remains available as historical actual data, while October 2027 is added to the future window.

Do not count the closed actual month as one of the next 12 forecast months. A model can display actuals and future forecasts side by side, but label which periods are included in each total.

Build the forecast from drivers

Use assumptions that explain operations rather than typing a percentage increase into every account. Choose drivers that are both important and practical to maintain.

Use Fixed vs. Variable Costs and Contribution Margin to establish cost behavior. A fixed cost can become a step cost when activity requires additional capacity.

A monthly update workflow

1. Close and validate actuals: import approved results from the month-end close and reconcile them to the source records.
2. Freeze the prior forecast: retain its date, horizon, assumptions, and version before making changes.
3. Explain the latest variance: separate volume, price, unit cost, timing, and errors using Variance Analysis.
4. Update remaining drivers: obtain evidence from sales, operations, people, procurement, and finance owners.
5. Add the new far-end period: preserve the selected future horizon and account for seasonality.
6. Recalculate linked statements and cash: verify that assumptions flow through consistently.
7. Review scenarios and decisions: identify resource, liquidity, and capacity consequences.
8. Approve and retain: document the review, changes, commitments, owners, and next update.

Do not automatically carry the latest month’s percentage variance across every future period. Establish whether the cause persists, reverses, or requires a different operating assumption.

Worked example: update the outlook after October

Original hypothetical example: this extends the single-product business in Budget vs. Actual Variance Analysis. The original October–December budget assumes 1,000 units per month, a $100 selling price, $60 variable cost per unit, and $25,000 fixed cost. Budgeted monthly operating profit is $15,000.

October closes at 1,100 units, a $98 price, $62 variable unit cost, and $26,000 fixed cost. Management then revises November and December using current orders and supported cost assumptions. The January forecast illustrates the next calendar year; it is not included in the October–December quarter total.

The price and cost improvements assumed for January need evidence, such as a supported pricing plan and supplier terms. The higher fixed cost needs an identified commitment and effective date. These figures are scenario inputs, not claims about what will happen.

Quarter outlook: October actual plus November and December forecasts produces $338,100 revenue, $213,900 variable cost, and $78,000 fixed cost, leaving $46,200 operating profit. The original quarter budget was $45,000 profit, so the latest estimate is $1,200 higher. October’s $1,400 shortfall is offset by forecast improvements of $400 in November and $2,200 in December.

This quarter view is an actual-plus-forecast estimate. The rolling next-12-month view after October still covers November 2026 through October 2027; the table shows only a short extract, not the entire rolling model.

Convert the operating forecast into cash timing

Revenue does not necessarily become cash in the month it is recorded. For a separate November cash illustration using the same revenue forecast, assume 70% of each month’s sales is collected in that month and 30% in the next month, with no losses or other collections.

November receipts: 70% × $112,700 + 30% × $107,800 = $111,230. Assume all November variable cost of $71,300 and fixed operating cost of $26,000 are cash payments in that month; no depreciation is included in those cost assumptions.

Cash falls $9,070 even though forecast operating profit is $15,400. Collection timing, capital spending, debt principal, interest, and tax payments explain why the measures differ. A complete model would also roll forward receivables, payables, assets, debt, and equity and verify the statement connections.

Use How to Build a 13-Week Cash Flow Forecast for detailed weekly timing and the cash-flow calculator to test short-term balances. The longer operating forecast and weekly cash view should reconcile over matching periods.

Build scenarios that support a decision

The following alternative November assumptions illustrate combined scenarios. They change several inputs together; this is different from a one-variable sensitivity test.

For each scenario, ask what management would do: delay discretionary spending, change capacity, renegotiate an input, or secure funding. Assign a trigger and owner. Do not assign probabilities or label a scenario “likely” without a supported basis.

A practical forecast structure

Keep inputs separate from calculated outputs. Mark actual and forecast periods clearly. Add checks for balance-sheet balance, cash rollforward, historical ties, sign conventions, period totals, and unsupported overrides. Preserve formulas when importing actuals.

An assumption log should show driver, old value, new value, effective date, source, reason, owner, and approval. This makes it possible to distinguish new information from model errors or unexplained optimism.

Measure forecast performance fairly

Compare actual results with the forecast version that existed before those results were known. Use the same horizon when comparing accuracy; a one-month-ahead forecast and a six-month-ahead forecast face different uncertainty.

Track both absolute error and signed bias. Opposite errors can cancel in a signed average while forecasts remain unreliable. Examine large misses by driver and improve assumptions rather than rewriting historical forecast values. See KPIs for cash-forecast error definitions.

Common mistakes

• Calling a year-end reforecast a constant-horizon rolling forecast.
• Overwriting the budget or prior forecast instead of retaining versions.
• Forecasting revenue without testing delivery capacity and collection timing.
• Treating all costs as variable or assuming fixed costs never step up.
• Using a profit forecast as a substitute for cash planning.
• Changing assumptions without an owner, evidence, or an effective date.
• Building more detail than the team can reliably maintain.

Frequently asked questions

Does a rolling forecast replace the budget?

It can complement the approved budget. The budget preserves a planning and accountability baseline; the forecast supplies the latest expected outcome. Choose the governance model deliberately.

How often should it be updated?

Often enough to support the decisions and respond to meaningful new evidence. Monthly updates are useful for many operating models; near-term cash or fast-changing conditions may require more frequent attention.

Can it be built in Excel?

Yes. Consistent definitions, linked drivers, version control, and review matter more than the software label. More complex organizations may need additional systems for consolidation and workflow.

Continue learning

Use Variance Analysis to explain changes, Financial Ratios to assess relationships, and Cash Flow Management to connect planning with liquidity. Return to Finance.

Sources and example notes

• Oracle: About Rolling Forecasts — continuously extending monthly or quarterly horizons.
• Oracle: Planning Revenue and Expense — driver and trend planning context.

The business scenario, calculations, cash bridge, scenarios, workflow, and model template are original educational illustrations. They are simplified planning assumptions, not promised results.