Annual plans are built on assumptions that are already aging by February. Deal cycles shift, headcount changes, and coverage gaps surface months before they show up in missed numbers. A rolling forecast replaces the annual plan as the operating document for revenue, keeping a constant forward horizon that updates as conditions change. This guide walks through how to design one from scratch.
Step 1: Choose Your Horizon and Granularity
A four-quarter rolling horizon is the standard for B2B SaaS revenue teams. It gives enough visibility for meaningful capacity and go-to-market decisions while staying close enough to actuals to remain credible.
Set the granularity by quarter for the outer periods and by month for the current and next quarter. Monthly granularity on the near term lets you catch problems before a quarter closes. Quarterly granularity on the outer periods is sufficient because you cannot reliably forecast individual deal timing that far out.
Your model structure should look like this:
| Period | Granularity | Primary Input |
|---|---|---|
| Current quarter | Monthly | Committed pipeline + closed |
| Next quarter | Monthly | Weighted pipeline |
| Q+2 | Quarterly | Weighted pipeline + capacity model |
| Q+3 | Quarterly | Capacity model + historical run rate |
Step 2: Build the Pipeline-Fed Input Layer
The near-term quarters of a rolling forecast should be fed directly from CRM pipeline, not spreadsheet estimates. Pull weighted pipeline by expected close date, segmented by the same categories you report against: new business, expansion, renewal.
For weighted pipeline to be useful here, your stage probabilities must reflect real historical close rates, not CRM defaults. If your historical data shows that deals in the Negotiation stage close at a different rate than your CRM's default, override the default. An inaccurate probability assignment in the pipeline layer compounds into a systematically wrong forecast.
Apply a forecast haircut to your weighted pipeline number for the current quarter. Weighted pipeline is almost always optimistic because it includes deals that will slip or die. Your haircut amount should be calibrated from historical forecast variance: the gap between what your weighted pipeline said at the start of prior quarters and what actually closed.
Step 3: Build the Capacity Layer for Outer Quarters
Quarters two through four rely more heavily on a capacity model than on visible pipeline. The capacity model answers: given our current and planned headcount, what revenue output should we expect per period?
Inputs to the capacity layer:
- Quota per fully-ramped rep, by segment - Current headcount by segment, with start dates for new hires - Ramp schedule (what percentage of full quota a rep produces in their first, second, and third quarters) - Historical attainment rate (not every rep hits quota; the capacity model should reflect actual attainment, not plan)
The output is a capacity-based revenue ceiling for each future quarter. If your weighted pipeline for Q+3 exceeds your capacity ceiling by a wide margin, that gap is likely sandbagging or stale pipeline. If it falls well short, you have a coverage problem to address now.
Step 4: Add Scenario Bands
A single-point forecast number gives false precision. A rolling forecast should present three scenarios for each future period: a conservative case, a base case, and an upside case. These are not arbitrary ranges. They should be driven by specific assumptions.
A practical approach:
- Conservative: pipeline held at current stage conversion rates, no deals advance stages, no new pipeline created - Base: pipeline advances at historical stage conversion rates, new pipeline created at current pace - Upside: pipeline advances at the top quartile of historical conversion, plus your expected new pipeline for the period
The spread between conservative and base cases gives leadership a real view of execution risk. If the conservative case is far below the base case, you have a pipeline concentration problem: a few large deals that may or may not close are carrying the number.
Step 5: Design the Governance Cadence
A rolling forecast without a review cadence becomes a spreadsheet that no one trusts. The cadence should be simple enough that it does not add meeting load on top of existing forecast calls.
A workable structure:
- Weekly: RevOps reviews pipeline movement and flags material changes to deal status or close dates. No meeting required. A short written summary distributed to the revenue leadership team. - Monthly: Revenue leadership reviews the updated rolling forecast, including scenario bands, coverage ratios, and any changes to capacity assumptions. This is the moment to adjust the outer-quarter numbers. - Quarterly: The monthly review for the final month of the quarter doubles as a handoff. The closing quarter becomes actuals, the outer quarter gets added, and assumptions for the new far quarter are set.
The monthly review is where a rolling forecast beats an annual plan. If a coverage gap appears in Q+2 during the monthly review, you have two to three months to respond before it becomes a miss.
Review revenue forecasting methodology alongside your rolling model to confirm your inputs are calibrated against the right historical period.
Common Mistakes
Treating the rolling forecast as a target, not a prediction. A forecast is a model of likely outcomes given current conditions. If leadership adjusts the forecast upward to match a target, the forecast loses its function. Keep the forecast honest. Address gaps through pipeline generation or capacity changes. Not separating new business from expansion. New business pipeline and expansion revenue from existing accounts have different lead times, conversion rates, and volatility. Blending them produces a forecast that is hard to act on. Model them separately. Locking outer-quarter assumptions too early. The further out the quarter, the more the forecast should rely on capacity and historical run rates rather than specific deals. If you are calling specific deals in Q+4, you are confusing pipeline visibility with forecast accuracy. Updating the forecast only when asked. The value of a rolling model comes from its continuity. If it only gets updated before a board meeting, it is still functioning like an annual plan.Frequently Asked Questions
What is a rolling forecast in sales?
A rolling forecast is a revenue model that always projects a fixed number of periods forward, regardless of the calendar year. When one period closes, the forecast adds a new period at the far end. This keeps the planning horizon constant and makes the forecast more responsive to current pipeline and market conditions than an annual plan locked in December.
How is a rolling forecast different from an annual plan?
An annual plan is fixed at a point in time and measures performance against assumptions made months earlier. A rolling forecast updates continuously as new data comes in. By the time an annual plan is six months old, its assumptions about pipeline coverage, deal cycles, and market conditions are often obsolete. A rolling forecast prevents the plan from going stale.
How often should a rolling forecast be updated?
Most revenue teams update their rolling forecast monthly, with a lighter weekly refresh on near-term pipeline. Quarterly updates are too infrequent to catch deal slippage or coverage gaps before they become misses. Weekly full rebuilds add overhead without meaningfully improving accuracy beyond the nearest quarter.
What data inputs does a rolling forecast need?
The core inputs are weighted pipeline by close date, historical stage conversion rates, average sales cycle length by segment, and current quota capacity. Secondary inputs include expansion revenue signals from existing accounts, churn risk from customer success data, and any known one-time items that will not recur.
Frequently Asked Questions
What is a rolling forecast in sales?
A rolling forecast is a revenue model that always projects a fixed number of periods forward, regardless of the calendar year. When one period closes, the forecast adds a new period at the far end. This keeps the planning horizon constant and makes the forecast more responsive to current pipeline and market conditions than an annual plan locked in December.
How is a rolling forecast different from an annual plan?
An annual plan is fixed at a point in time and measures performance against assumptions made months earlier. A rolling forecast updates continuously as new data comes in. By the time an annual plan is six months old, its assumptions about pipeline coverage, deal cycles, and market conditions are often obsolete. A rolling forecast prevents the plan from going stale.
How often should a rolling forecast be updated?
Most revenue teams update their rolling forecast monthly, with a lighter weekly refresh on near-term pipeline. Quarterly updates are too infrequent to catch deal slippage or coverage gaps before they become misses. Weekly full rebuilds add overhead without meaningfully improving accuracy beyond the nearest quarter.
What data inputs does a rolling forecast need?
The core inputs are weighted pipeline by close date, historical stage conversion rates, average sales cycle length by segment, and current quota capacity. Secondary inputs include expansion revenue signals from existing accounts, churn risk from customer success data, and any known one-time items that will not recur.
See how ORM turns these insights into action
ORM builds custom revenue forecast models for B2B SaaS companies. Not dashboards. Prescriptive analytics that tell you what to do next.
Schedule a Demo