Engineering consultancies: why utilisation targets hide the real constraint

Engineering consultancies run on utilisation. Every engineer has a chargeable target, the target is reported monthly, and performance against it drives resourcing decisions and often remuneration. The discipline is real and it is one of the reasons the model works.

It also produces a recurring puzzle. A firm reports utilisation in the low eighties across the technical team, which by industry convention is healthy, and margin that does not reflect it. The numbers are not being manipulated. They are measuring something narrower than the leadership team believes.

What utilisation does not tell you

Utilisation records that time was booked to a job. It says nothing about whether the time was recoverable, whether it was within the fee, or whether it was necessary.

An engineer who spends eleven hours on a task quoted at six is fully utilised. So is one redoing work because a design assumption changed after the client approved it, or producing a deliverable to a standard the fee never contemplated. The firm records high utilisation and writes off the difference at invoicing, where it appears as a recovery problem rather than a delivery one.

Write-offs are where the real signal lives, and in many firms they are handled quietly by the project lead at the end of a job, discussed with no one, and never aggregated. Nobody sees the pattern because nobody is looking at it as a pattern.

Four structural drains that utilisation conceals

Scope drift accepted at technical level

Clients ask engineers for things. Engineers, being professionals who want to solve the problem in front of them, do them. The request is technically modest and commercially unfunded, and it is agreed between two technical people with no reference to the fee proposal. By project completion the delivered scope and the sold scope have diverged materially, and the variation conversation happens after the work is done, when it is hardest to win.

Review capacity concentrated in a few people

Deliverables require sign-off from a senior engineer or an RPEQ or equivalent, and in most mid-sized firms that sits with a small number of individuals. Those people are also the ones winning work and managing clients. Review becomes the bottleneck: projects wait, deadlines compress, and junior work is checked in a rush at the end rather than guided at the start, which increases rework.

Bidding that is not informed by delivery data

Fee proposals are typically built from a previous similar job, adjusted by judgement. If the firm does not track actual effort against fee by job type, that judgement is being informed by what was quoted last time rather than what it cost. Underpricing perpetuates itself, and the firm wins more of the work it makes least money on. This is the same dynamic that afflicts accounting practices, where the professional habit of absorbing overruns hides the pricing error.

Project management treated as an add-on

In many consultancies the project manager is a senior engineer doing it alongside technical delivery. Programme, budget tracking and client communication get the hours left over after the technical work, which means they are done reactively. Problems are discovered when they arrive rather than when they become likely.

A fully utilised firm with thin margin is not short of work. It is selling hours at one price and delivering them at another.

The metrics that show the actual constraint

Utilisation is worth keeping. It needs company.

Effort against fee, by project and by job type, reported while the job is live rather than at completion. Write-off rate by project lead and by client, aggregated so the pattern is visible. Review turnaround time, which quantifies the bottleneck everyone can feel. And variation capture rate: what proportion of out-of-scope work was identified, quoted and agreed before it was performed.

Those four numbers change the conversation from how busy people are to where the capacity is going. That is the distinction between reporting and governance: metrics that describe activity, versus metrics that support a decision.

Where firms usually start

The highest-return first move is generally variation control, because it is a process change rather than a systems project. Agree what constitutes a scope change, give engineers a fast and unembarrassing way to flag one, and make sure a commercially accountable person sees it within a day. Firms that do this consistently recover a surprising amount of revenue that was previously donated.

After that, look at review capacity, because it constrains both quality and throughput and it worsens as the firm grows. Widening the pool of people who can review, and moving their involvement earlier in the job, reduces rework at the same time as it removes the queue. Neither of these requires more engineers, which is usually the conclusion a firm reaches when it looks only at utilisation.