Billable hours: how many you really have, and the rate they need to earn

Every services business sells hours, whatever it calls them. A fixed-price project is a bet on how many hours it will take. A retainer is a block of hours with a monthly wrapper. Even value-priced work is delivered by people whose time you pay for by the hour, every hour, sold or not.

So the billable hour is the unit your whole business runs on. Most owners I talk to know their rate card. Far fewer know how many hours their team bills in a year, and almost none know the rate those hours need to earn for the business to hit its target margin. This article works through all three numbers. It is one of the two pillars of the billing hub, alongside retainer pricing.

What a billable hour is

A billable hour is an hour of work that you are entitled to invoice under the agreement with the client. The definition has two parts, and the second matters as much as the first. The hour has to be spent on the client’s work, and the contract has to let you charge for it.

The gray areas cause the arguments. Project management is the classic one. Clients sometimes object to paying for it, and agencies sometimes hide it inside inflated production hours. Both are worse than a plain line in the statement of work that says project management is billed at the same rate as everything else and typically runs 10% to 15% of project hours.

There is also a third category that matters more than either column: hours worked on client projects that never reach an invoice. I will come back to those, because they are where the money goes.

How many billable hours a person really has

Start with the paid year and subtract.

That leaves 1,400 billable hours, a realistic average across a delivery team that mixes juniors, seniors and leads. That is about 27 hours in a 40-hour week, averaged over the year.

Owners react badly to this number. It feels low. Then they check their own timesheet data and find it is about right, or generous. A plan built on 1,800 billable hours per person is built on a year in which nobody takes a holiday, attends a meeting or waits for a client to send content.

The number varies by role, and I give the ranges I consider sane in the article on utilization rate. For planning a whole team, use these as a starting point and replace them with your own data as soon as you have it.

From hours to required rate

Here is the calculation every owner should be able to do on one sheet of paper. I will use one example agency throughout.

Eight delivery people. Loaded cost of $150,000 per head. Each bills 1,400 hours a year. The owners want a 25% net margin.

“Loaded cost” needs a definition, because it is where most rate calculations go wrong. It is the full cost of running the business divided by the number of billable people. Salary is only part of it.

Now the rate.

Two numbers come out. The break-even rate is $107.14: below that, an hour of work loses money. The required rate is $142.86: that is what every billed hour must earn, on average, for the business to make its 25%.

Note the margin step. Required revenue is cost divided by one minus the margin. Adding 25% to cost gives $1,500,000, which is a 20% margin. This mistake is common and it costs five points of margin before the year has started.

Put your own numbers in.

Bill rate and realized rate

The agency in the example has a rate card of $140 an hour. That is $2.86 under the required rate, which looks close enough.

It is nowhere near close, because the rate card is the rate you ask for. The realized rate is what you collect, divided by the hours you worked. Here is the same agency’s year.

The rate card says $140. The business earned $123.21. The gap to the required rate is $142.86 less $123.21, or $19.65 an hour. Across 11,200 hours that is $220,000 of missing revenue.

Nobody in this agency did anything dramatic. The team was busy all year. Utilization looked healthy. The rate card was competitive. The profit still came in at less than half the plan, and the whole shortfall sits in four lines of leakage that no dashboard showed.

This is why I tell owners to enter the rate they collect when they use the calculator. The rate card flatters you. The realized rate is the one your bank account agrees with.

Where billable hours leak

Each of the four leaks has a different fix.

Discounts

Retainer discounts and volume deals are legitimate. They are also cumulative, and nobody adds them up. A 10% retainer discount on a third of your hours takes about 3% off your realized rate across the whole business. Give discounts only in exchange for commitment, and cap them. The schedule I use is in the retainer pricing pillar.

Fixed-price overruns

This is the largest leak in most project shops. You quoted 400 hours, the job took 520, and the extra 120 hours earned nothing. At $140 that is $16,800 on one project. The fix is upstream: honest contingency, a tight scope, and a decision about whether the project should have been fixed price at all. The scoring sheet and calculator for that decision are in fixed price vs time and materials.

Write-offs

A client disputes an invoice and you credit 20 hours to keep the peace. Sometimes that is the right call. It becomes a pattern when invoices arrive without context. An itemized invoice with a short summary of what was delivered gets disputed far less than a single line reading “Development, 86 hours”. If disputes have turned into non-payment, see what to do when a client is not paying.

Unbilled extras

The “quick favor” is the smallest leak per incident and the most habit-forming. A developer spends 40 minutes on something the client mentioned in a call and never logs it against a billable code. Multiply by eight people and fifty weeks. The fix is partly a rule, that all client time is logged against the client, and partly a process for anything outside the agreed scope. That process is a change request, and the larger pattern it prevents is scope creep.

Tracking hours without making everyone miserable

You cannot manage realized rate without time data, and time tracking is the least loved process in any agency. Some rules that make it work.

Track everything, including fixed-price and internal work. If only billable time is logged, you cannot calculate utilization, and you cannot see what the non-billable hours are being spent on. Internal time needs codes too: sales, training, admin, bench.

Log daily. A timesheet filled in on Friday afternoon is a work of fiction. Daily entry takes three minutes and is accurate to within fifteen. Weekly entry takes twenty minutes and is accurate to within a day.

Keep the code list short. Client, project, and a handful of task types. Forty task codes produce worse data than six, because people pick the first plausible one.

Use sensible increments. Fifteen minutes is the norm for agency work. Six-minute increments come from law firms and produce resentment and noise in a creative or engineering team.

Never punish honest data. The moment someone is criticized for logging 9 hours on a task estimated at 5, everyone learns to log 5. You then lose the only information that could have fixed your estimates.

Review weekly, by project. Hours used against budget, for every active project, every Monday. A project that is 60% through its hours and 40% through its scope is a conversation to have this week.

Raising billable hours and raising rate: which lever pays

When profit is short, the instinct is to push for more billable hours. Compare the levers on the example agency.

Getting everyone from 1,400 to 1,500 hours is a 7% increase in workload. It adds about $98,500 and it comes out of training time, slack and goodwill. Closing the leaks adds $188,000 with no extra hours worked. The rate lever is larger than the hours lever in almost every agency I have looked at, and it is the one owners reach for last.

There is a ceiling on the hours lever as well. Past a certain point, more billable hours means no time for sales, no time to learn, and no capacity to absorb a project that runs long. I explain why a fully booked team is a warning sign in the utilization rate article.

Billable targets: use them carefully

Many agencies give each person an annual or weekly billable target. Targets are useful for planning and risky as an incentive.

A target of 30 billable hours a week tells a developer what the business needs from them. That is fair and clarifying. A bonus tied to hitting 30 hours tells the same developer to make sure 30 hours are logged, and the cheapest way to do that is to round up, pad, and avoid the non-billable work the agency also needs: mentoring, code review for other teams, helping with an estimate.

My preference is to set targets by role, publish them, review them monthly at team level, and keep individual pay out of it. Bonuses, where they exist, work better tied to project margin or company profit.

People also cannot bill hours that have not been sold. When a developer’s billable hours fall, the cause is usually an empty pipeline or a stalled client. Holding the individual accountable for a sales problem is unfair and they know it. The planning side of this, including what to do with people between projects, is covered in bench management.

Billable hours under fixed price, retainers and productized work

Billable hours are easiest to see under time and materials, where each one becomes an invoice line. The other models hide them, and the hours still decide whether you made money.

For a fixed-price project sold at $60,000, the question after delivery is simple. If it took 400 hours, the realized rate was $150. If it took 520, it was $115.38, which in the example agency is $8 above break-even and well under the required rate. Without timesheets, both projects look identical in the accounts: $60,000 of revenue each.

A side-by-side view of how each model treats hours and risk is in agency pricing models.

When hours stop being the right unit to sell

Billing by the hour has a built-in conflict. The better and faster you get, the less you earn for the same result. For years this was a slow effect, because experience makes people perhaps 20% or 30% faster over a career.

Tooling has changed the size of the effect. When a task that took ten hours takes four with AI assistance, an hourly agency has a choice between passing the entire saving to the client and quietly billing ten. The first shrinks revenue. The second is dishonest and eventually discovered. The better route is to change the unit: price the outcome through fixed fees, deliverables retainers or productized services, so that efficiency improves your margin. I go into the broader shift in is software engineering dead.

Even then, you keep counting hours internally. The hour stops being what the client buys. It remains what you pay for, and the required-rate calculation above applies in exactly the same way. Only the label on the invoice changes.

A monthly routine that takes one hour

Run these five numbers on the first Monday of each month.

1. Billable hours by person, against the target for their role.
2. Utilization for the team: billable hours divided by available hours.
3. Realized rate: revenue invoiced for the month divided by client hours worked.
4. The gap: realized rate less required rate, multiplied by hours. This is the month’s surplus or shortfall against plan, in dollars.
5. Leakage by cause: discounts, overruns, write-offs, unbilled time. Whichever is largest gets attention this month.

In the example agency, a typical month has 933 client hours. At a realized rate of $123.21 the month brings in about $115,000 against a requirement of $133,300. The routine would have shown an $18,000 shortfall in January, with fixed-price overruns as the largest cause. That is eleven months of warning the owners did not have.

Once you know your required rate, the next question is how to price above it. That path, from cost floor to market rate to value, is laid out in how to price consulting services.