Revenue forecasting for project businesses

A project business does not forecast revenue from the pipeline, but from its order book: which contracts have been awarded, when the work will be carried out and at what pace it will be completed. The forecast therefore consists of the schedule of running and awarded projects, supplemented by service and maintenance, and only then by tenders not yet awarded. The biggest risks lie not in winning work, but in delays, capacity, and additional work that is carried out but never invoiced.

Why project businesses forecast differently

In a business that sells subscriptions or products, a won deal becomes revenue within a few weeks. In a construction, installation, engineering or implementation business there can be a year between award and completion. Revenue comes in over that period as the work progresses, and that pace is rarely even.

Three things make it different:

1. Revenue follows execution, not sales. A EUR 1,200,000 contract awarded in March often produces nothing in March.
2. Capacity is the constraint. You can sell more than you can deliver. Then revenue slips, however large the order book.
3. Revenue and invoicing diverge. Payment schedules follow the contract, not progress. The difference sits on the balance sheet as work in progress.

A pipeline forecast from the CRM therefore says little about next quarter's revenue in a project business. Why that is true in general is explained in why pipeline is not revenue. With projects the gap is even larger.

The building blocks

1. Running projects: what is still outstanding

For each running project you need: the total contract value including approved variations, what has already been completed, and the schedule for the remainder. You spread the remaining value over the months according to that schedule. This is the most certain part of your forecast, but only as certain as the schedule.

2. Awarded projects not yet started

Contracts that are agreed but have not begun. Here the start date is the biggest uncertainty. Permits, preparation on the client's side, materials with long lead times and the completion of other projects determine when you actually start.

3. Service, maintenance and small jobs

Many project businesses have a recurring layer: maintenance contracts, call-out services, small jobs that come back every month. You forecast these from history, as described in forecasting from historical revenue. This layer is often more stable than the project work and deserves more attention than it usually gets.

4. Quotes and tenders

Only last: work that has not yet been awarded. For each quote, a probability of award and, more importantly, a realistic start date if it is awarded. For the current quarter this layer contributes little in most project businesses. For the next six months it is decisive.

How revenue is spread over a project

Most projects do not follow a straight line. The beginning is preparation and mobilisation, the middle is full production, the end is finishing and handover. This is often called an S-curve.

Worked example: suppose you are an installation business with a EUR 1,200,000 project running for eight months. From comparable earlier projects you know that the first two months together account for around 10 percent of production, months 3 to 6 together for 70 percent (17.5 percent a month) and the last two months for 20 percent.

The payment schedule in the contract is: 30 percent at start, 30 percent after month 4, 30 percent after month 7 and 10 percent on completion.

Look at months 3 to 5, the coming quarter:

• Revenue (production): 3 x 17.5 percent = 52.5 percent of EUR 1,200,000 = EUR 630,000.
• Invoicing: only the second instalment falls in this quarter, EUR 360,000.

In this quarter EUR 270,000 more revenue is earned than invoiced. That is not a problem as long as it is deliberate. It becomes a problem when instalments structurally lag behind progress, because then you are financing your client's work.

And if the start of this project slips by two months, a large part of that EUR 630,000 falls into the next quarter. The forecast for the full year barely changes; the one for the quarter changes enormously.

Where the forecast structurally goes wrong

The schedule is optimistic. Project managers plan how things should go, not how they usually go. Compare planned and actual duration for your completed projects. If they took 15 percent longer on average, that correction belongs in your forecast.

Award is not start. In many sectors there are weeks to months between award and start. If revenue starts on the award date in your forecast, you will be structurally too early.

Capacity is double-counted. Each project on its own is feasible. Together they require more fitters, site managers or engineers in the same quarter than you have. Put the sum of the schedules next to your capacity. Where it exceeds it, revenue slips.

Additional work is carried out but not invoiced. A change agreed on site or with the client is carried out, but never makes it onto a variation form and therefore never onto the invoice. That is not a timing difference, but revenue that is gone. It is one of the biggest leaks in the sector, worked out in revenue leakage in project businesses.

Work in progress is not monitored. If nobody tracks per project how much has been completed and how much invoiced, the difference grows unnoticed. At the end of the project it turns out that part of it can no longer be charged.

Scenarios are essential in project businesses

Because a few large projects determine revenue, a single forecast is too fragile. Work with scenarios around the events that really matter: the start of the largest new project slips a quarter, two tenders are awarded at once and your capacity falls short, a running project stops for a month. How to model them is covered in scenario forecasting for B2B.

The most useful scenario for a project business is often not the pessimistic one, but the one in which everything arrives at once. That shows whether your capacity can cope, and whether your cash flow can carry the pre-financing.

What the sector adds

In construction there are also instalments tied to milestones and disputes over variations and omissions at completion. How Revenue Intelligence works there in practice is covered in Revenue Intelligence for construction. In installation and technical services, the balance between projects and service is often the key: a stable service layer absorbs fluctuations in project work.

Step by step

1. List all running and awarded projects with total value, approved variations, completed to date and planned end date.
2. Spread the remaining value of each project over the months, using a realistic curve based on earlier projects.
3. Correct for historical delay. How much longer did your projects take on average than planned?
4. Add up the schedule per month and set it against your capacity. Move what does not fit.
5. Add service and maintenance based on history.
6. Add quotes with probability of award and realistic start date, shown separately.
7. Compare completed with invoiced for each project. Every growing difference is a question for the project manager.
8. Each month, compare additional work carried out with additional work invoiced.

How this fits into the wider discipline of forecasting is covered in the complete guide to revenue forecasting.

Frequently asked questions

Can a project business forecast with its CRM?

For quotes and tenders, yes. For next quarter's revenue, no: that comes from the project administration and the schedule. Many project businesses use their ERP or a project system alongside the accounting package for this.

How do you handle projects with a lot of additional work?

Forecast approved variations as part of the contract value. Keep additional work that has not yet been formally approved separate, and make sure it gets on paper as quickly as possible. Only then is it revenue you can invoice.

What is good accuracy for a project forecast?

On an annual basis, a project business with a well-filled order book can be reasonably accurate. Per quarter the spread is larger, because schedule changes feed straight through. Measure your own deviation over a few quarters and use it as your margin.

Should you forecast revenue or invoicing?

Both, separately. Revenue for managing results and capacity, invoicing for cash planning. The difference between the two, work in progress, is a management figure in its own right.