Revenue Forecasting for a Professional Services Firm
Services revenue is not booked, it is earned, and the gap between those two words is where most forecasts go wrong.
A product business forecasts what it will sell. A services business has to forecast what it will sell, when the work will actually be performed, and how much of the performed work will convert to revenue. Each of those is a separate estimate with its own error, and combining them without acknowledging that is how a forecast comes to be confidently wrong.
The way through is to build the forecast in layers, each with its own confidence, rather than as one number.
The three layers
- Contracted backlog: work already signed and not yet delivered. The most reliable layer, and it still requires a view on timing, because a signed engagement that starts two months late moves the revenue into a different period.
- Extensions and renewals: work with an existing client that has not been signed but has a history. Weight it by the firm's actual renewal rate rather than by optimism, and treat a renewal in negotiation differently from one nobody has yet raised.
- New pipeline: weight by stage using historical conversion, and be explicit that the further out it sits, the more the estimate reflects capacity to sell rather than demand.
Fee models change the arithmetic
On time and materials work, revenue follows effort, so the forecast is a resource schedule multiplied by rates, and its main error is the schedule. On fixed-fee work, revenue is typically recognised as the work is performed, which means the forecast depends on an estimate of progress, and an engagement running over budget earns no more revenue for the extra effort while consuming the cost.
That asymmetry is worth stating plainly to anyone reading a services forecast, because it produces a counter-intuitive effect: on fixed-fee work, an engagement that is going badly can look normal in the revenue line and terrible in the margin line at the same time. A forecast that shows only revenue conceals exactly the problem the reader most needs to see.
The errors that recur
- Forecasting the sale rather than the delivery. Work signed in March and delivered in the second half produces very little revenue in the first, and pipeline-driven forecasts routinely miss this.
- Ignoring holiday and seasonality. Available hours fall sharply in some months, and a forecast built on a flat monthly capacity overstates the summer and the year end in most markets.
- Treating a renewal as certain because it always renewed. Concentration risk is real, and the year a large renewal does not happen is the year the forecast was least prepared for it.
- Forgetting the ramp. New joiners are not fully productive on day one, and a plan that counts them at full utilisation from their start date overstates capacity by a predictable amount.
- Confusing revenue with cash. A profitable quarter with slow collections can leave a firm unable to pay itself, so the cash forecast needs its own view of billing timing and days outstanding.
Reforecasting on a rhythm
A forecast produced annually and defended monthly is a political document. A forecast reproduced monthly, with the previous version kept, is a management tool, because the interesting information is in the movement rather than in the number.
Keep the variance analysis short and causal: what moved, and was it timing, scope, price, or loss. Those four categories cover almost every movement, and distinguishing them matters because they call for different responses. Timing is a planning problem, scope is a delivery conversation, price is a commercial one, and loss is a pipeline one.
What to report alongside it
A revenue forecast alone is not enough to run a services business. Three companions make it useful: the resource plan it implies, so anyone can see whether the revenue is deliverable; the margin at forecast rates, so a growing revenue line built on discounting is visible; and the backlog coverage, which is the proportion of the next period's forecast that is already contracted.
Backlog coverage is the single most informative number for a services business and the least often reported. A firm entering a quarter with eighty percent of its forecast contracted is in a different position from one entering with forty, and no amount of pipeline commentary conveys that as clearly.