Eight hours of work, two and a half hours of pay

Because the hours you work and the hours you are paid for are two different quantities, and the ratio between them is far worse than almost anyone assumes. Paid hours = available hours × utilisation × realisation × collection. Four terms, multiplied together. In legal practice the product of the last three is about 31%. An eight-hour day pays for two and a half hours.

The rate — the number on your proposal — is a fifth variable. It is also the only one most people ever touch.

The identity that decides what lands in your account

It is worth writing out, because almost nobody does:

Paid hours = Available hours × Utilisation × Realisation × Collection

• Available hours — the hours you are at work at all.
• Utilisation — the share of those hours recorded against a client.
• Realisation — the share of recorded hours that reach an invoice, after write-downs, courtesy reductions and the hour you quietly did not charge for because the job took longer than you said it would.
• Collection — the share of invoiced money that actually arrives.

Revenue is then paid hours × rate. Five variables in total. Four of them govern the hours. One governs the price. The price is the one that gets discussed.

Eight hours in, two and a half hours out

Clio publishes benchmarks for legal practice in its Legal Trends Report. The 2025 edition, with data running through October 2025, puts utilisation at 38%, realisation at 88% and collection at 93%. Those are three ordinary-looking numbers. Multiply them.

8 hours × 38% = 3.04 hours recorded against a client.
3.04 × 88% = 2.68 hours invoiced.
2.68 × 93% = 2.49 hours collected.

2.5 hours of an eight-hour day reach the bank account, once utilisation, realisation and collection have each taken their share. A capture rate of 31%.

The other five and a half hours were real. They happened, they were worked, they are gone.

The same source reports lockup — the delay between doing work and holding the money for it. Realisation lockup 43 days. Collection lockup 32 days. Total lockup 93 days. The three are reported separately and the total is not the sum of the two components, so I will not pretend they add; what matters is the headline figure. Ninety-three days is a quarter of a year between the work and the payment for it.

And this is the improved version of the picture. Clio's 2017 Legal Trends Report recorded 2.3 billable hours a day at 29% utilisation, and 1.6 paid hours a day. Eight years later the profession has moved utilisation from 29% to 38% and paid hours from 1.6 to roughly 2.5 — half again as much. That is real improvement, achieved by an industry with billing software, timekeepers and partners whose compensation depends on the number.

It arrives at two and a half hours.

Now run it backwards

This is the part that stops people, and it is worth doing slowly.

To collect CHF 200,000 at a rate of CHF 200 an hour, you need 1,000 paid hours.
At a capture rate of 31%, 1,000 paid hours requires 1,000 ÷ 0.31 = 3,226 hours at work.
3,226 ÷ 52 weeks = 62 hours a week.

Sixty-two hours a week, every week of the year, with no holiday, no illness, no gap between clients and no week in which anything goes wrong.

Effort was never the variable. If the capture rate is 31%, there is no number of hours that reaches the target — the arithmetic runs out before the week does.

The practical translation of all this is a single multiplication. Your true earnings per hour of working life are roughly your headline rate × 0.31.

CHF 200 an hour becomes CHF 62. CHF 350 an hour becomes CHF 108.50. €103 — the average hourly rate reported by 5,412 freelancers across the DACH region for 2026 — becomes about €32.

The 31% comes from North American legal practice, and applying it to a Zurich consultancy is an illustration rather than a measurement. I would rather say that than let the number do more work than it can carry. But the direction is not in dispute, because the same shape turns up everywhere anyone has bothered to measure it.

The gap is structural, and it has a size

SPI Research publishes an annual benchmark of professional services firms. The 2025 edition, its eighteenth, covers 2024 data from 403 firms.

Billable utilisation by firm size. SPI Research, 2025 Professional Services Maturity Benchmark, 2024 data, n=403 firms.

Firm size	Billable utilisation
Under 10 employees	64.3%
10–30 employees	65.4%
31–100 employees	69.2%
101–300 employees	72.4%
301–700 employees	73.7%
Over 700 employees	78.3%
All firms, 2024	68.9% (71.4% in 2020)

Fourteen percentage points separate the smallest firms from the largest. That is the structural fact underneath this entire subject. A firm of 700 people is not more diligent than you are. It is organised so that most of its people are never asked to do two jobs in the same hour.

Clio's 2024 report found the same thing from the other end: larger firms run utilisation rates about 15% higher than solo firms.

The RIBA Business Benchmarking Report 2023, published on 6 December 2023, measured UK architecture practices and found the pattern in its purest form. Partners, directors and sole principals spend 46% of their time on billable work. Associates, 64%. Architects, technologists and assistants, over 70% — rising above 80% in practices of 100 staff or more. The most senior and most expensive person in the practice is the least billable one. Overall practice profit margin in that report had compressed to 2%.

The number people report and the number that gets measured

Here are two facts that are usually printed in different places.

The Freelancer-Kompass 2026 — freelancermap's survey of 5,412 freelancers across the DACH region, fielded between 17 November 2025 and 8 February 2026 — reports that 12% of total working time is non-billable, about 5 hours of a 42-hour week. That is the self-reported figure.

SPI measures firms under ten people at 64.3% billable, which is 35.7% not. RIBA measures sole principals at 46% billable, which is 54% not.

Three populations, three methods, and I am not going to pretend they are comparable. What they share is a direction: every measured figure is dramatically worse than the self-reported one. RIBA supplies the reason in a single line — only 60% of Chartered Practices record billable time at all. You cannot report a number you have never computed. What you report instead is an impression, and impressions are generous.

The composition of that unpaid time is worth noting. Asked what fills it, 58% of freelancers named customer acquisition, 35% accounting, 25% bureaucracy, 14% networking and 2% self-marketing. The shares exceed 100% because respondents named more than one. Selling is the largest single item, which is a structural problem of its own.

The weeks that appear in nobody's utilisation figure

Everything above measures the inside of a working week. It says nothing about the weeks with no work in them.

IPSE's Freelancer Confidence Index found UK freelancers spent an average of 3.5 weeks per quarter with no work at all in the second quarter of 2024, and 3.4 weeks in the third. The organisation does not publish a sample size for those figures, and that matters — treat them as indicative rather than precise. Even so: 3.4 weeks of a 13-week quarter is 26% of the year with no billable work available, before a single minute of the non-billable time inside the working weeks is counted.

The freelancermap data points the same way. 24% of respondents had fewer than 50 project days in the past year. 43% have no secured project utilisation for the coming months.

Which term is worth attacking, and why it is almost never the rate

The identity is worth a second reading, because the four terms are not the same kind of thing.

Realisation and collection describe money that has already been earned and already been agreed. 88% × 93% = 81.8%, which means 18.2% of the work you recorded never becomes money in your hands. Recovering it requires nobody's permission. Nobody has to accept a new price. It is an administrative act that has never been assigned to anyone.

Utilisation sits at 38% in that legal benchmark and 64.3% in firms under ten people. Whatever your own figure is, it has room above it. The rate does not have the same room, because the rate is bounded by what a market will pay and moving it requires other people to agree.

Then compare the two levers on equal terms. A 20% increase in rate produces 20% more revenue. Moving utilisation from 38% to 46% — eight percentage points, which is 0.64 of an hour a day — produces 46 ÷ 38 = 21% more paid hours, and therefore 21% more revenue. The same result. One of them requires a negotiation with every client you have. The other requires a calendar.

This is not an argument that your rate is correct. In my experience it usually is not, and whether you are undercharging is a separate question with its own arithmetic. It is an argument about order. Fixing the rate moves one term of five, and it is rarely the term that is broken.

What this means in practice

The four terms are recoverable from two documents you already own: your calendar and your invoice ledger. Available hours are in the calendar. Utilisation is the share of those hours you can attach to a named client. Realisation is invoiced hours divided by recorded hours, which requires you to have recorded something you did not bill — the write-downs are the point, and they are the entries most likely to be missing. Collection is money received divided by money invoiced, with the ageing on top.

The reconstruction takes about a fortnight of unglamorous work for a year of history, and what it produces is a single number — your capture rate — plus a decomposition telling you which of the four terms is furthest from where it could be. Until that number exists, every decision about price, hours or capacity is being taken against an impression.

What I usually find is that people know their rate to the franc and cannot state their utilisation to the nearest ten percentage points. That asymmetry is not carelessness. It is that one of the two numbers is printed on every document they send, and the other is printed nowhere at all.

What this cannot tell you from the outside

Every figure above comes from a population that is not yours. Clio measures North American legal practice. SPI measures professional services firms large enough to run time systems. RIBA measures UK architecture. freelancermap measures DACH freelancers, weighted towards IT and consulting. IPSE publishes no sample size. None of them is a benchmark for a single Swiss practice, and anyone who tells you otherwise is selling something.

What none of them can tell you is your own four terms — and here is the awkward part. Your financial statements cannot tell you either. The accounts report collected revenue, which is available hours × utilisation × realisation × collection × rate, already multiplied out into one number. The statement reports the product and destroys the factors. You cannot recover four unknowns from one equation, which is why the answer to "where did my year go" has never once been found in a profit and loss account.

Nobody in your professional circle has been engaged to produce the factors. The fiduciary reports the year that closed. The bank looks at the total. The person best placed to reconstruct the four terms is you, from your own calendar, and it is the one piece of work nobody has ever asked you for. The numbers that decide a business are almost never the ones its owner has been given.