UK engineering consultancies face a rare pipeline at the same times as a structural skills shortage. The firms that thrive will double revenue per engineer, not headcount. Alison Murtagh, who leads UK marketing at Projectworks - a professional services automation platform for engineering consultancies – explains more

Engineering consultancies in the UK are holding two truths at once.

The pipeline is the strongest it has looked in years, with a 10-year infrastructure plan on the table and projects such as Sizewell C, the Transpennine Route Upgrade and the Lower Thames Crossing moving from headline to delivery.

But the people needed to deliver it are structurally scarce, with costs rising and the gap widest for firms outside the major cities.

Many have heard that AI resolves this, however far fewer can point to where it actually shows up in margin, capacity or the bottom line. Closing that gap starts with arithmetic rather than tools. If you can't hire your way to the pipeline in front of you, the only route through is delivering more with the engineers you already have.

AI doesn't change the fundamentals of running an engineering consultancy

AI doesn't retire the basics of running a consultancy. Work still has to be won, delivered at margin and kept clear of admin creep. Projects on publicly funded programmes are expecting tighter scopes, faster turnaround, and cleaner reporting, and that accountability is filtering from the boardroom to the project site. What AI changes is the speed at which a disciplined firm can compound its strengths. Firms with tight commercial fundamentals will multiply them. Firms without will simply generate chaos faster.

AI disruption in professional services is arriving on a two-year cycle

Each general-purpose technology wave has arrived faster than the last. Electrification took decades to reshape industry, the internet took years and cloud, mobile, and big data compressed the cycle further still. Generative AI reached mainstream use in under two years, with advanced AI following in less. For professional services, disruption now arrives on roughly a two-year cycle. The planning horizon for wait-and-see has effectively disappeared.

What factory electrification teaches engineering consultancies about AI

Research by Azeem Azhar and Nathan Warren at The Exponential View offers a useful parallel for AI adoption: electricity did not transform factories simply by being installed. Lighting made workplaces safer but the real productivity gains came only when Ford, for example, rebuilt the factory floor around individual electric motors and the flow of work.

That is the lesson for UK engineering consultancies. A chatbot in a second window is the lightbulb. The firms that win are those mapping how work is scoped, resourced, delivered and billed, using AI to run or support repeatable tasks, and moving engineers onto the judgment work that only they can do.

First the individual speeds up, then the workflow, then the firm.

Revenue per engineer is the metric that matters

If AI is genuinely making a consultancy more productive, one number should move: Revenue per engineer.

Tool adoption and hours saved are activity measures. Revenue per engineer is an outcome measure, and it's honest about whether AI has changed your delivery capacity or just your software bill. It also reframes the skills shortage. A 30-person consultancy that lifts revenue per engineer by a third has effectively added ten engineers it never had to recruit, in a market where those ten may not exist to hire.

Push further, with connected data, live visibility, and agents handling resourcing and project admin, and doubling revenue per engineer inside two years is a credible target. A target, to be clear, not a guarantee, and directors should treat it as a hypothesis their own numbers will test.

Five metrics tell you if it's working

Measurement is where most AI initiatives quietly fail, so anchor to the fundamentals.

1. Gross margin shows whether the firm is healthy at all
2. Utilisation belongs at 70-80%, high enough to be productive, low enough to hold a bench that's ready for new work
3. Book-to-bill reveals whether you're growing or shrinking, and catches the quiet leakage of scoped work that's delivered but never invoiced
4. The share of revenue from advisory and design-led work, versus body-shopping engineers by the hour, shows whether you're competing on judgment or capacity
5. Tech utilisation, the share of each engineer's week lost to admin that software built for engineering consultancies could carry, is the number AI should visibly shrink first

The window in front of UK engineering is rare. Political clarity, public investment, and a technology inflection do not often arrive together. The line between growth and chaos is thin. Which side will your firm land on?

• Alison Murtagh leads UK marketing at Projectworks, a professional services automation platform purpose-built for engineering consultancies. She spends her time talking to firms across the sector about the practical realities of adopting AI — where it moves the needle on margin and delivery capacity, and where it's just noise.