Ask a services firm how it is doing and it will often answer with one number: utilization. Billable hours as a share of available hours, firm-wide, for the month. A good month is a number the partners are pleased with. A bad month starts a conversation about whether people are busy enough.

The number is useful for one thing, which is noticing that something has changed. It is useless for the next question, which is where. A firm-wide average has no address. It does not tell you whether the leak is in the senior people, the junior people, one client, one kind of work or the gap between what was logged and what was invoiced.

Why the average misleads.

An average adds up opposing movements and reports the sum. A firm whose junior consultants are fully booked and whose senior consultants spend a third of their week on proposals and internal meetings will report an acceptable average. Both facts are true, and the average cannot tell you the second one exists.

Worse, the common calculation divides billable hours by hours worked, and "billable" is a label the person applied when they logged the time. Hours labelled billable are not the same as hours billed. Between the two sit write-downs the partner made before the invoice, hours the client disputed, hours on a fixed-fee project that were billable in the timesheet and absorbed in the contract, and hours that were never invoiced because nobody got to them. The label says one thing and the cash says another.

We described how three systems compute the number three ways in the utilization you cannot see. Agreeing on one definition is the first step. The second step is to stop reporting it as a single figure.

Four buckets.

Take every hour logged in the month and put it in exactly one of four buckets.

Billable and billed. Hours that were logged as billable, appeared on an invoice at the rate they were meant to carry, and that the client paid or will pay. This is the revenue-bearing bucket.

Billable but written down. Hours logged as billable that appeared on an invoice at a reduced rate or a reduced quantity, or were removed before invoicing. The write-down may have been right. The point is that it is visible as a bucket, with a reason code.

Billable but not invoiced. Hours logged as billable that appear on no invoice yet. Some are timing, because the invoice is next week. Others are the hours that will never be billed, either because the project's budget ran out or because nobody has looked. Age this bucket. Hours older than the billing cycle are a leak, not a timing difference.

Non-billable by choice. Hours the firm decided not to bill: proposals, training, internal projects, a client relationship it was building. These are an investment and should be a planned figure, not a residue.

The four buckets add up to the hours logged. The share of billable and billed is the number worth watching, and the other three are where you look when it moves.

Read it by role, not by firm.

The split becomes useful when it is cut by role, project and client, one at a time.

By role, you will usually find the leak concentrated. Senior people have a large non-billable-by-choice bucket, which is expected, and a surprising billable-not-invoiced one, because their time is the last to be coded and the first to be forgotten. Junior people have almost nothing in not-invoiced and a larger written-down share, because their time is what clients challenge.

By project, you find the fixed-fee work whose hours were billable on paper and absorbed in practice. Those are the projects where the estimate was wrong, and the written-down bucket is the evidence.

By client, you find the account that takes a share of senior time out of proportion to its revenue, or the one whose invoices are always trimmed. These are conversations about price, scope or whether to keep the account, and they cannot start from an average.

What it takes to produce.

The split needs three links that a spreadsheet-run firm usually lacks. Each logged hour must carry a project, a role and a billable label. Each invoice line must point back to the hours it covers, so that billed hours can be matched with logged ones. And write-downs need a reason code, even a short one.

The second link is the one most firms have not built. Without it, billed hours are known only in total, from the invoice, and the difference between logged and billed is a single unexplained number. With it, each hour has a state. The same discipline appears in what breaks between the proposal, the project and the invoice: the joins are where the information is lost.

An illustration, with made-up numbers.

Imagine a twelve-person firm that logs 1,600 hours in a month. The dashboard says 70 percent billable, which looks fine. Split it: 1,000 hours billed, 90 written down, 30 billable but not invoiced and still ageing, and the remaining 480 non-billable by choice, a third of which is one senior person's proposal work.

Billed hours are 62.5 percent of the hours logged, not 70. The gap between the label and the cash is eight percentage points, and the split says where it sits: ninety hours that someone reduced and thirty that nobody invoiced. On a day rate those hundred and twenty hours are a measurable amount, and each bucket has an owner who can look at its own lines this afternoon.

Now cut the same month by role. Suppose the write-downs sit almost entirely with two junior consultants on one fixed-fee client, and the not-invoiced hours are all one partner's. That is two conversations, one about the client's scope and one about the partner's billing habit, and neither would have started from the average.

How often to look, and who owns what.

Weekly for not-invoiced, because it ages, and monthly for the rest. The owner of not-invoiced is whoever raises invoices. The owner of written-down is the partner who approved the reduction, who should be able to say why. The owner of non-billable by choice is the managing partner, who set the target for it.

Each owner gets a number they control. The firm-wide average, which nobody controlled, is kept as a headline and treated as the output of the four owners' work.

The first thing to try.

Do not wait for a system. Take last month's timesheet export and last month's invoices, and match them by project and week in a spreadsheet. It will be rough, and it will still show the four buckets in outline, with a number against each. The roughness itself is useful, because the lines you cannot match are the missing link between a logged hour and an invoice, and that is the link to build.

Where we come in.

We connect the time entries, the project records and the invoice lines so that every hour has a state, and we build the four-bucket view by role, project and client for the people who own each bucket. Discovery is no-risk: we run the split on last quarter's hours and invoices and show where the leak sits, and you pay only if you go ahead. The build carries a guaranteed estimate, so if we estimate low, we absorb the difference.

An average tells you how the firm is doing. The four buckets tell you where to stand to change it.

Find out what one connected Zoho system would change in the way you run.

In a no-risk discovery we look at your CRM, finance and operations systems and who owns each part, and show what a connected system would do differently. You pay only if you proceed. Or see how we approach it.

Book the no-risk discovery

More on the same problem:

Where we're not the right answer

We'll tell you if we're a fit. If we're not, we'll tell you that too.

  • Your current stack works and nobody wants to change it
  • You want licences resold at a discount and nothing else
  • Your internal team owns the operating model and is not handing it over
  • You want hours of configuration work and nothing run for you: that is on our services pages, and it is not a managed solution
How we start

Most of our best clients come to us with a feeling, not a plan.

"Something isn't working." "We're outgrowing our tools." "We're afraid to make the wrong move." No-Risk Discovery is a short, practical conversation that gets you clarity before you commit to anything big. We'll tell you if we're a fit. If we're not, we'll tell you that too.

Talk to us

Have a question about this?

Say who you are and ask. A person replies within one business day.

A new client, or thinking about becoming one
What are you looking for?
An agentic solution (AI OS) is licensed and sold with managed services. An integrated solution is built for a professional services fee. A Service is professional work, done by the hour, the month or the project. Not sure yet is a fine answer.

Agentic solutions (AI OS)

Integrated solutions

Services

Platforms You Use or Are Considering

Your details are never shared or sold.