A services business (service business, in the singular, or professional services firm when the work is knowledge work) earns its revenue by selling the time and expertise of people rather than copies of something it made once. Consultancies, agencies, managed service providers, law and accountancy practices, IT support companies and training firms are all services businesses. The defining fact is that revenue is bounded by three things: how many billable people the firm has, what it charges for them, and how much of their time it sells, which is its utilisation (utilization, in the US spelling). Everything that makes a services business different from a product company follows from that, and most founder advice on the internet is written for the product company.
The formula: people times rate times utilisation
Revenue in a services business is, near enough, billable people multiplied by the rate they are sold at multiplied by their utilisation, which is the share of their available time that is actually billed. Three levers, no more:
- People. Add a billable person and you add capacity. You also add their salary from the day they start, weeks or months before they are fully billed.
- Rate. Raise the rate card and every billed day is worth more. Rate is the lever with the least cost attached and the most fear.
- Utilisation. Sell more of the time you already pay for. It is the fastest lever and it has a ceiling: above a certain point the people who are busiest are the ones who leave.
An invented example, for method only: twelve billable people, sold at an average of £900 a day, billed for 65 per cent of 220 available days each, produce about £1.54m of revenue a year. Take utilisation to 75 per cent and the same twelve people produce about £1.78m without a single hire. Add two people at the same rate and utilisation and you add about £297,000, minus two salaries from their first day. Those figures are invented to show the arithmetic and are not a benchmark for any firm.
The number that pulls the three levers together is revenue per billable person, which is the one I would put on the wall of a services board pack before any other.
Why the maths is different from a product company
| Product company | Services business |
|---|---|
| Make once, sell many times; marginal cost of the next sale is close to zero | Every sale is delivered again by a person; marginal cost is a salary |
| Revenue can grow without headcount | Revenue tracks headcount; to grow you hire, and you hire before the revenue arrives |
| Gross margin is a property of the product | Gross margin is a property of rate, utilisation and salary, and moves every month |
| Cash arrives on subscription or at purchase | Payroll goes out monthly; invoices are paid on thirty to ninety day terms behind the work |
| Valued on recurring revenue and growth rate | Valued on profit, discounted for founder dependence and rewarded for contracted, recurring work |
| The product is the asset | The people are the asset, and they can resign |
The consequence founders feel first is cash. A services business pays for its capacity in advance, every month, and is paid for it in arrears. Growth makes that worse, not better, because each new hire is a cost this month and revenue in three. The consequence founders feel last is valuation: a buyer of a services business is buying the people, the contracts and the founder’s willingness to stay, and will discount anything that depends on the founder selling it again next year.
Why founder advice written for software misleads
Most of what a new founder reads was written by and for software companies, because that is where the money and the writing have been for twenty years. Applied to a services business, it misleads in specific ways:
- “Focus on growth, margin comes later.” In software, scale improves margin. In services, growth is hiring, and hiring ahead of revenue lowers margin and drains cash. Margin has to be managed every month, alongside growth, not after it.
- “Track ARR and MRR.” Those definitions assume a subscription. A services firm’s recurring revenue can be counted four different ways from the same ledger, and a buyer will count it the strictest way. Borrow the vocabulary and you will believe a number nobody else does.
- “Build once, sell everywhere.” A services firm can productise, and should, but a productised service is still delivered by a person each time. The marginal cost never reaches zero.
- “Raise money to grow faster.” Growth capital in a services business funds a shape of company, and accelerates a wrong one just as effectively as a right one. We took growth investment in 2018 and planned the following year in billable heads; the spreadsheet row that decided what kind of company we built was the hiring row, not the product row.
- “Your customers are your moat.” Your people are. Utilisation above the level they can sustain is a churn forecast.
None of this means the software playbooks are wrong. They are answering a different question, and a services founder needs to know which parts transfer. A consultancy, an agency and an MSP each need a version of this translation, and they differ in what they sell and how they are valued.
What this site is about
I started the business that became DevOpsGroup, a services company helping other businesses build and run software, with my co-founder Steve in 2013, a services company that helped other businesses build and run their software. It was sold for over $30m in 2021. Through all of it, I was looking for writing by someone who had built a business like ours, and found almost none: founder-to-founder advice for people who sell skilled time. This site, and the newsletter it grew from, exist to be that writing. The post that explains why is the place to start. Everything on the homepage is organised around the maths on this page: rate, utilisation, recurring revenue and what a buyer will eventually count.