The Professional Services Utilization Crisis: What 68.9% Means

Service Performance Insight publishes their professional services benchmark every year. The 2025 edition put the industry’s average billable utilization at 68.9%, against a target most teams set for themselves around 72%¹. That is the headline number that has been circulating in operator forums since the report landed.

The number is real. It is the wrong thing to focus on.

I have spent enough time inside services teams to know that utilization, taken as a single industry-wide figure, hides more than it reveals. The teams sitting at 68.9% are not all suffering from the same condition. Some of them are dealing with a soft pipeline. Some are over-hired against a forecast that did not land. Some are losing margin to scoping mistakes that show up as bench time after the fact. Some are tracking the wrong thing entirely. The interventions for those four conditions look completely different. The number alone tells a practice head none of that.

This piece is about what the 68.9% figure actually measures, where the gap between booked and effective utilization comes from, why utilization hides delivery margin, and the four operating moves that move the needle.

What utilization actually measures

Billable utilization is the percentage of an operator’s available capacity that is recorded against billable client work. The exact formula varies by team, but the standard SPI definition treats it as billable hours over standard available hours, where standard available hours is something close to a working week minus holidays, training, and other non-billable allocations¹.

That is the booked figure. It is what shows up in dashboards. It is the figure compared across teams in industry benchmarks like SPI. And it is the figure that operators tend to optimize against when nothing else is in their dashboard.

Booked utilization is not the same as effective utilization. Two teams sitting at the same booked utilization can have very different operating realities once you account for what those billable hours actually delivered. One team’s 72% billable hours produces realised margin in the high twenties. Another team’s 72% billable hours produces realised margin in the low single digits because half those hours were spent on out-of-scope work the team is absorbing. Both teams report 72%. Only one is healthy.

The 68.9% number is a lagging indicator of operator health. It is not the disease.

The booked-versus-effective gap

The gap between booked utilization and effective utilization is where most services teams are actually losing.

A team’s booked utilization can be made up of any combination of:

• Billable hours against scope, margin holding.
• Billable hours against scope, margin compressing because effort exceeded estimate.
• Hours billed at concession or against a stretched scope, margin negative.
• Hours allocated to internal investment work that gets coded billable for utilization reporting purposes.

The booked figure does not distinguish those four. Most dashboards do not either. A team can have a clean-looking 72% utilization rate that is structurally thirty points lower in effective terms because the engagement scope has drifted under the team’s feet for six months.

This is the operator tax that the 68.9% figure is hinting at without naming. The gap is not entirely a pipeline problem. A meaningful share of it is engagement drift: scope that has expanded against the original quote without a corresponding price adjustment, deliverable inflation against the original specification, and out-of-scope work the team is absorbing because raising it as a change would jeopardise the relationship. Most teams I work with do not have a clean instrument to see that drift in real time. The drift shows up as a margin surprise three or six months later, by which point the engagement is in trouble.

For the deeper operator read on this specific failure mode, see preventing scope creep in professional services.

Why utilization hides delivery margin

Practice heads sometimes ask me why their utilization looks healthy and their margin looks bad. The answer almost always sits in the gap above. The dashboard is tracking a metric that goes up when consultants log billable hours, regardless of what those billable hours actually achieve.

Three patterns show up consistently across the teams where this is true.

The first is that the engagement layer is fragmented. Discovery notes live in one system, the quote lives in another, the resource plan in a third, the actuals in a fourth. There is no single object that holds the engagement’s state from sale through delivery. Without that, the team cannot see, at a given week, whether the engagement is still on the trajectory it was sold on.

The second is that scope is captured as text and tracked as text. The original SOW is a Word document. Scope changes are emails. Steerco-level scope adjustments are slide decks. None of those are queryable. None of them roll up into a structured view of how the actual scope diverged from the quoted scope. The drift is invisible to the dashboard.

The third is that delivery actuals do not feed back into the next quote. The team’s own history is not legible to the team’s next estimator. The same misestimate gets paid for again on the next engagement of the same shape. The booked utilization looks fine. The effective utilization, the part that funds the practice, keeps eroding.

These three patterns are why a number like 68.9% can be both true and almost useless. The figure is a symptom. The operator gap underneath it is the disease. If you have read AI Won’t Save Your Services Business, the structural argument here will be familiar: it is the same diagnosis from a different angle. For the sibling read on the operator paradox of high utilization against poor outcomes, see the utilization paradox.

The four moves that actually move utilization

There are four operating moves that, in my notebook, show up consistently in teams that have moved their utilization in a structural rather than a cyclical way.

1. Make the engagement layer the unit of truth

One structured object per engagement, holding state across discovery, quoting, resource plan, delivery, and closeout. Not a folder. Not a tag in a PSA. A working object the team’s actual workflow runs against. Once it exists, the booked-versus-effective gap becomes visible because the engagement holds both numbers.

2. Convert scope from text to structure

The SOW is a document for legal. The team’s working scope needs to be a structured set of artifacts that the engagement object holds: deliverable specifications, complexity drivers, defined acceptance, named assumptions, named risks. When a scope change happens, the artifact changes. The drift becomes a number, not an anecdote. For the working artifact most teams use to make this transition, see the SOW template and the companion piece on why the SOW process is broken at most teams.

3. Close the loop between actuals and the next quote

Delivery actuals against quoted effort, on a per-engagement basis, fed back into the team’s complexity drivers and into the similar-engagement archive. The team’s own history becomes legible to the team’s next estimator. This is the institutional memory work I described in the end of tribal knowledge.

4. Make resource decisions against capacity, not against utilization

A resource decision optimized against utilization puts a consultant on whatever billable thing is available. A resource decision optimized against capacity asks whether the consultant is the right shape for the engagement, against actual planned delivery rather than dashboard slots. The first move maximises the booked figure. The second moves the effective figure. The companion working artifact for this is the utilization dashboard template, which separates booked from effective in a way most off-the-shelf reporting does not.

Why none of this is a pipeline answer

Operators sometimes read pieces like this and conclude that the answer is more pipeline. More demand, the logic runs, and utilization goes up.

Pipeline matters. It is not the answer to the gap between booked and effective utilization. A team with weak structural controls and strong pipeline will absorb the same operator drift at higher volume and end up with worse margin. The 2025 SPI data shows industry profit margins at decade lows even as some segments of the industry have not had a corresponding pipeline collapse². The shape of the problem is operator, not commercial.

The teams that are emerging from this period stronger are the ones who took the 68.9% figure seriously enough to look at what was underneath it, and then made one of the four structural moves above. The number does not move because someone redoubles their utilization push. It moves because the operator gap closed.