A practical guide to moving from static annual budgets to continuous rolling forecasts. Covers why rolling forecasts outperform static budgets, how to design one, selecting the right drivers, establishing update cadence, and avoiding the most common pitfalls.
Static annual budgets are obsolete the moment they are approved. Market conditions shift, customer behaviour changes, and strategic priorities evolve — yet the budget remains frozen at a point-in-time snapshot from three months ago. Rolling forecasts solve this fundamental problem.
A company that sets its annual budget in November is making assumptions about December of the following year — 13 months away. By Q2, many of those assumptions are wrong. The budget becomes a fiction that finance teams measure against out of obligation rather than insight.
A rolling forecast always looks 12-18 months ahead, regardless of where you are in the fiscal year. When January closes, you add a new month (or quarter) at the end of the horizon. This means you always have a forward-looking view of the same length. Research from the Association for Financial Professionals shows that companies using rolling forecasts report 25% faster decision-making and 10% higher forecast accuracy compared to those using static budgets alone.
You are ready for a rolling forecast when your annual budget process takes more than six weeks, when mid-year re-forecasts feel like starting from scratch, or when leadership routinely asks "what do we actually think will happen?" rather than "what did we budget?". Most UK companies with £5M+ revenue benefit from the shift.
A well-designed rolling forecast balances granularity with maintainability. The goal is enough detail to be useful, but not so much that updates become burdensome.
The standard rolling forecast horizon is 12-18 months. The near-term (next 3 months) should be at monthly granularity with high confidence. The medium-term (4-12 months) can use monthly or quarterly periods with moderate confidence. The outer months (13-18 months) are quarterly and directional. This graduated approach keeps the model manageable.
Forecast revenue at the product or segment level, not individual customer level. Forecast headcount costs at the department level, with individual position detail for the next quarter only. Forecast operating expenses at the cost category level (travel, software, professional services), not individual line items. The 80/20 rule applies: 20% of your line items drive 80% of the variance.
Build your rolling forecast with scenario capability from day one. The base case is your working forecast. Maintain at least one upside and one downside scenario that share the same structure but use different assumptions. This allows rapid scenario switching without rebuilding the model.
The best rolling forecasts are driver-based: they start with operational metrics and translate them into financial outcomes. This makes the forecast more intuitive, easier to update, and better connected to business reality.
Start by mapping the five to ten operational drivers that most influence your financial results. For a SaaS business, these might include: new logo pipeline (£), win rate (%), average contract value (£), net revenue retention (%), and new hires per quarter. For a services business: billable headcount, utilisation rate (%), average day rate (£), and project win rate (%).
Structure your model in three layers. The input layer contains assumptions: "We expect 50 marketing-qualified leads per month at a 20% conversion rate". The calculation layer applies formulas: "50 leads x 20% conversion x £25,000 ACV = £250,000 new ARR per month". The output layer shows the P&L impact. This separation makes it easy to update assumptions without touching formulas.
Once your drivers are defined, rank them by impact. Use a simple sensitivity table: "If win rate drops from 20% to 15%, revenue falls by £X". This tells you where to focus your forecasting effort. Drivers with high impact and high uncertainty deserve the most attention and the most frequent updates.
A rolling forecast only works if it is actually updated. That requires a clear cadence and genuine buy-in from the people who own the assumptions.
Most companies update their rolling forecast monthly, shortly after the monthly close. A typical timeline: actuals are finalised by working day 5, the FP&A team updates the forecast model by working day 7, department heads review and adjust their assumptions by working day 10, and the consolidated forecast is published by working day 12. Some fast-moving businesses update weekly for cash forecasting.
The biggest risk to a rolling forecast is that it becomes an FP&A-only exercise. Department heads must own their assumptions. Give each budget owner a simple input form — not a 50-tab spreadsheet — where they update three to five key drivers each month. Make it clear that the forecast is a planning tool, not a performance contract. People game budgets; they improve forecasts.
Present the rolling forecast alongside actuals and the original budget. Show three numbers for each line item: budget, current forecast, and actuals-to-date. Highlight where the forecast diverges from budget and explain why. Over time, the forecast should converge with actuals as you approach each period, giving leadership confidence in its reliability.
Rolling forecasts are conceptually simple but operationally challenging. Here are the most common mistakes and how to avoid them.
Some teams build rolling forecasts with the same granularity as their annual budget — hundreds of line items, individual cost codes, customer-level revenue. This makes monthly updates impossibly slow. The fix: reduce granularity. A rolling forecast with 30 well-chosen line items updated monthly is infinitely more valuable than one with 300 line items updated quarterly.
If nobody owns the revenue growth assumption, nobody updates it. Assign every key driver to a named individual. The VP of Sales owns pipeline conversion. The Head of Engineering owns the hiring plan. The FP&A team owns the model structure, not the assumptions.
A forecast should reflect your best estimate of what will happen, not what you want to happen. If the sales team knows their forecast will become their target, they will sandbag. Separate the forecast (what we expect) from the plan (what we are aiming for). Report both, but do not conflate them.
If updating the forecast requires copying data between five spreadsheets, it will not survive the first busy month. Automate the data flow: actuals should pull automatically from your accounting system, and driver inputs should feed directly into the model. This is where dedicated FP&A software like Grove FP makes a material difference.
Related Templates
Move beyond the static annual budget with a rolling forecast that extends 12 months ahead at all times. As each month closes, actuals replace the forecast and a new month is added to the end. Includes variance tracking against the original budget.
A 13-week (one quarter) cash flow forecast tracking weekly inflows and outflows. Covers customer receipts, supplier payments, payroll, tax, and discretionary spend. Shows opening balance, net cash movement, and closing balance for each week.
A driver-based revenue model that builds your top-line forecast from core assumptions — new customers, average revenue per user, retention rates, and expansion. Supports subscription (ARR/MRR), transactional, and recurring revenue models.
Related Tools
Work out how many months of cash runway your business has remaining. Enter your current cash balance and monthly burn rate to see your runway end date and plan accordingly.
Measure your revenue growth rate between any two periods. Calculate period-over-period growth percentage and compound annual growth rate (CAGR) to benchmark your business performance.
Grove FP makes it easy to implement the processes described in this guide. Build budgets, run forecasts, and produce board-ready reports in one platform.
FAQ
The standard horizon is 12-18 months. The near-term (next 3 months) should be monthly with high confidence, the medium-term (4-12 months) monthly or quarterly, and the outer months (13-18) quarterly and directional.
Most companies update monthly, shortly after the monthly close. Actuals are finalised by working day 5, and the consolidated forecast is published by working day 12. Some fast-moving businesses update weekly for cash forecasting.
Yes, and most companies do. The annual budget serves as a baseline for performance measurement and incentive compensation. The rolling forecast provides the current best estimate. Report both side by side for maximum context.
At minimum, you need a structured model (not ad-hoc spreadsheets) with clear input and output layers. As complexity grows, dedicated FP&A software like Grove FP provides automated actuals integration, driver-based modelling, and real-time collaboration that make monthly updates sustainable.
Budgeting, forecasting, and workforce planning in one platform. No credit card required.