Revenue Forecasting for Service Businesses
Here's how most service businesses forecast revenue:
The founder looks at the open pipeline, does some mental maths about which deals they think will close, adds some expected new business, and produces a number that sounds reasonable.
This isn't forecasting. It's optimism with a spreadsheet.
Optimism-based forecasts cause serious problems: hiring decisions made too early, cash flow shortfalls in the month the forecast didn't materialise, capacity investments that can't be unwound.
A real forecast — one you can make decisions with confidence from — is built from data, historical conversion rates, and explicit assumptions. Here's how to build it.
The Three-Bucket Revenue Forecast Model
Divide your revenue projection into three buckets:
Bucket 1: Committed Revenue (High Confidence, 90-100%)
This is revenue that is contracted or highly likely to continue. It includes:
- Signed retainer agreements for the forecast period
- Renewal clients in the final stage of renewal confirmation
- Milestone-based projects where the milestone has been reached and invoicing is confirmed
Committed revenue is your floor. If you hit nothing else in the forecast period, this is what you'll have.
Bucket 2: Pipeline Revenue (Medium Confidence, 40-70%)
This is revenue from deals currently in your active pipeline that have a realistic probability of closing within the forecast period.
The key is probability weighting by stage:
| Pipeline Stage | Historical Close Rate | Probability Weight | |---|---|---| | Discovery call completed, proposal imminent | 55% | × 0.55 | | Proposal sent, active discussion | 38% | × 0.38 | | Proposal sent, awaiting response | 22% | × 0.22 | | Qualified lead, discovery call scheduled | 15% | × 0.15 |
Sum all opportunity values multiplied by their probability weight.
Example: £50,000 proposal in active discussion = £50,000 × 0.38 = £19,000 probability-weighted contribution.
This weighted sum is your pipeline revenue contribution — more accurate than simply summing all open opportunity values.
Bucket 3: New Business Revenue (Lower Confidence, 10-30%)
This is revenue from new leads not yet in your pipeline. It's the hardest to forecast accurately but should still be modelled — based on your historical lead velocity and conversion rates.
If your historical lead velocity is 15 new qualified leads per month and your full-funnel conversion rate is 12%, you can expect 1.8 new clients per month × average deal value.
For a 90-day forecast: 5.4 expected new clients × average deal value. Apply a 70-80% confidence factor to account for timing variability.
The Three-Bucket Model in Practice
For a service business at the start of a quarter:
Committed revenue:
- 8 active retainer clients × £3,200/month × 3 months: £76,800
Pipeline revenue (probability-weighted):
- £25,000 proposal in active discussion × 38%: £9,500
- £18,000 proposal awaiting response × 22%: £3,960
- £12,000 qualified lead, call scheduled × 15%: £1,800
- Total pipeline contribution: £15,260
New business revenue:
- Expected 4.5 new clients × £3,200 × 80% confidence: £11,520
Total 90-day forecast: £103,580
This is your planning number. Not a hope — a data-based projection with explicit assumptions you can interrogate and stress-test.
Stress-Testing the Forecast
A good forecast isn't just a number — it's a range.
Base case: The three-bucket model as calculated above.
Conservative case: Pipeline conversion rates -30% (deals close slower or not at all), new business -40%. In the example: £76,800 + £10,680 + £6,912 = £94,392
Upside case: Pipeline conversion rates +20%, new business +25%. In the example: £76,800 + £18,312 + £14,400 = £109,512
The range (£94,392 to £109,512) tells you the bandwidth you're operating in. Your committed revenue (£76,800) is your worst-case floor. Make sure your cost structure is viable at that floor.
What Good Forecasting Requires
To run this model, you need:
- Pipeline stages with historical close rates — tracked in your CRM for at least 6 months
- Open opportunities with values — every deal needs a £ value entered
- Expected close dates — every deal needs a timeline
- Lead velocity data — how many new qualified leads enter per month
- Committed revenue visibility — current retainer clients and their renewal dates
If any of these are missing, your first investment is in data infrastructure, not forecasting models.
The Weekly Forecast Review
Revenue forecasting is not a quarterly event. The best businesses do a 5-minute version every week:
- How did this week's revenue compare to the weekly equivalent of the quarterly forecast?
- Did any deals close faster or slower than expected?
- Did any new opportunities enter the pipeline that should be added to the model?
- Are there any committed revenue risks (retainer clients who've shown dissatisfaction)?
This weekly cadence means you're never surprised at the end of a quarter. You see the quarter forming in real time.
Book a free audit call and we'll build your specific revenue forecasting model using your actual pipeline data and historical conversion rates.