Utilisation Is a Bad KPI.

Tracking utilisation makes sense.

As agency leaders, we're effectively paying for 38 hours per week of an FTEs time.

That is our single biggest expense line in the P&L. So obviously we want to know that time is being spent producing as much revenue as possible.

But tracking a metric and KPI-ing your team on it are two very different things.

Why utilisation is a poor KPI for do-ers

Most people on your team don't control the inputs.

They've got the work they've got. If they're doing all of it, that's basically their maximum utilisation short of doing billable work for free. So if you KPI them on a number they can't meaningfully influence, what are you actually optimising for?

If you do that, you're probably incentivising them to log more billable hours you're not getting paid for.

Charlie Munger said it best:

Show me the incentive and I'll show you the outcome [let's substitute 'outcome' for 'behaviour' here].

If you set 80% utilisation as a KPI and you're not winning enough new business to keep the team busy, they'll find a way to hit 80% anyway. That means overservicing clients, marking time as billable that probably isn't, or just spending longer on work that should've taken half the time.

Both outcomes are bad. But the second one is really bad, because now you've got no accurate picture of what's happening in the business, and you've actively incentivised the behaviour that caused it.

Where utilisation is useful

That's not to say you shouldn't track it. You absolutely should, just in the right context.

For team leads, it's a useful diagnostic. If your team is telling you they're flat out, but utilisation is sitting at 60%, that's worth investigating. Are they spending too much time on non-billable work? Is meeting prep or client comms not being marked as billable? Is the data even accurate?

The utilisation metric isn't the problem. It's a lagging indicator pointing you toward whatever the actual problem is.

It's also useful as a point-in-time calibration shared with the broader team, rather than used as a target.

If utilisation is way lower than it should be, you can have an honest conversation about efficiency or pipeline. If it's sitting at 110%, that explains why everyone feels like they're drowning, and gives you a framework for what to do about it: hire, invest in tools, or just tell the team it's a busy patch and it will pass.

Running the business means you have a huge amount of context the team doesn't - and sharing that context helps people understand the bigger picture of the business they're working in.

The suite of metrics problem

The deeper issue with utilisation isn't really about utilisation. It's about relying on any single metric. Even the good ones fall apart in isolation.

Revenue per employee looks great until you add in the fact that your team is paid way above average (or they're all offshore and comparatively cheaper).

Project profitability looks great until you realise your team is at 110% to achieve it.

No single number tells the whole story.

A decent agency dashboard would include things like overall profitability, contribution margin, revenue per employee, project and client-level profitability, service line performance against target bands, client retention/growth, and flags for any clients being materially over or underserviced. Utilisation sits somewhere in that mix.

What the team should actually be KPI'd on

Behaviours and outcomes they control.

For account leads, that means things like client profitability, retention, and growth. For everyone, it includes what I refer to as 'Operational Excellence' - accurately tracking the work that feeds metrics like Utilisation, using PM systems, getting invoices out on time etc etc.

Utilisation as a KPI gets the team focused on a single number which isn't actually that useful.

What you actually want them focused on is the work, the clients, and the accuracy of the data that helps you run the business well.

Those aren't the same thing.