An IT services firm forecasts revenue it can explain
Revenue forecasts were built bottom-up in spreadsheets by each delivery head, and changed every week without a reason. A driver-based rolling forecast tied the number to headcount, utilisation and rates.
An illustrative scenario. The company is fictional and the figures show how the method works; they are not a client result.
The situation
An IT services firm delivers projects and managed services to clients in several countries. Every month, delivery heads sent revenue forecasts in their own spreadsheets. Finance consolidated them, and leadership questioned them. The forecast moved every week, and nobody could say why.
What we looked at
- Three years of revenue by client, project and contract type
- Headcount, skills and allocation data
- Time sheets and utilisation
- Billing rates, rate cards and contract terms
- The sales pipeline, with stage and probability
How the forecast was rebuilt
Revenue was forecast from its drivers, not as a single number per account:
| Contract type | Drivers |
|---|---|
| Time and materials | Billable headcount, utilisation, billing rate, working days |
| Fixed price | Milestones, percentage complete, change requests |
| Managed services | Contracted fee, volume bands, service credits |
| Pipeline | Deal value, expected start, probability by stage |
Vittora produced a baseline forecast from history and current allocations. Delivery heads adjusted it where they knew better, and every adjustment carried a reason.
What we found along the way
- Unbilled time. Hours logged against billable projects but never invoiced, mostly from late time sheets.
- Rate gaps. People billed at lower rate-card levels than their current grade.
- Bench drift. Utilisation falling in some skills while others were short, visible only once the data was joined.
What changed
- A rolling forecast refreshed each month, eighteen months ahead.
- A bridge every month showing why the forecast moved: actuals, drivers or judgement.
- Accuracy published by business unit, so the forecast improves over time.
- Unbilled time chased weekly, before it becomes a write-off.
Read The rolling forecast that explains itself, or see forecasting and FP&A.