The AI Pivot

Why UK Private Equity Is Falling Back In Love With the Physical Economy

An EMC perspective on how AI is rewriting the UK mid-market deal playbook.

June 2026

Spend a morning in Mayfair in 2026 and you’ll hear a conversation that would have been unthinkable five years ago. Partners at UK mid-market houses who spent a decade chasing SaaS businesses are now talking about ductwork contractors, accountancy roll-ups and HVAC networks. The hottest decks crossing our desks at EMC are no longer opening with “AI-powered.” They’re opening with “AI-proofed.”

Nobody in the UK lower mid-market is quite ready to call the end of the software cycle. But the appetite of capital allocators is clearly shifting, and it’s showing up in live processes right across the world.

The SaaS thesis starts to wobble

For fifteen years, SaaS was the closest thing private equity had to a sure bet. Recurring revenue, fat gross margins and predictable retention. Fund managers loved it. Valuations reflected that love.

Generative AI has cracked three of those pillars in quick succession. The cost of building decent software has collapsed – an AI-native team can now ship in a quarter what used to take a year, which is ruinous for any incumbent whose valuation rested on the assumption it was hard to displace. Switching costs are coming down too, because AI-native competitors are structurally better at ingesting a customer’s data and migrating them off legacy tools. And Finance Directors who happily signed annual renewals in 2022 are now kicking off procurement reviews with a single question: what if we just didn’t renew?

Not every software business is struggling. The deals that still clear at full multiples tend to share a profile: proprietary data, regulated or networked workflows, genuine two-sided dynamics – not a thin AI wrapper bolted onto a mediocre product. That kind of software story still commands real conviction. Much of the rest of the category, frankly, is finding it harder to make the numbers work.

Rediscovering the physical economy

While software has been losing its halo, a the “boring” kind of business is getting a proper second look from UK sponsors: facilities services, M&E contractors, specialty logistics, veterinary and dental groups, testing and inspection, and industrial services of every flavour.

These are the businesses PE spent years finding unfashionable. Vans, engineers, apprentices, dispatchers, twenty-four-hour call-outs — in the unkind shorthand of a previous era, the “unscalable” parts of the British economy. That cliché hasn’t aged well. Three things have changed the picture.

One: the moats are real. A regulated services business with engineers, accreditations and physical presence cannot be disrupted by a GPT wrapper. The constraints that used to make these businesses dull – licensing, labour, logistics – are exactly the constraints that now make them defensible against AI.

Two: AI is now the operating leverage. A services roll-up today can put AI to work across dispatch, call-centre automation, quoting, route planning and back-office functions. The margin story is no longer about cutting heads – it’s about getting more out of the ones you have. That translates to real EBITDA expansion without the brand damage of a traditional cost-out programme, which is a story UK funds can now underwrite with genuine confidence.

Three: the demographic backdrop is substantial. The UK’s owner-operator generation is retiring. Tens of thousands of healthy, cash-generative businesses will change hands over the next decade – and that supply of assets is arriving at exactly the moment the thesis for buying them has sharpened.

People-led businesses get a second look

The same logic is breathing new life into what used to be considered among the riskier buyout categories: partner-led professional services. Accounting firms, law firms, architecturual practices and engineering consultancies were historically avoided because “the asset walks out the door every night.”

AI is quietly changing that. Tacit knowledge can be captured in systems. First drafts get produced in minutes. The junior-heavy leverage pyramid can be re-engineered. Put those three things together and the economics of a professional services firm start to look less like a partnership and more like a platform. The UK accounting consolidation wave – Sumer, Cooper Parry, Xeinadin, S&W – is the most visible sign, and law firm platforms are following close behind. AI isn’t replacing these professionals; it’s making each of them more productive – and that’s exactly the kind of repeatable uplift a sponsor can build a thesis around.

What it means for business owners

For owners thinking about a sale or investment over the next two to three years, this shift is worth taking seriously.

If you run a software business, the question is no longer simply whether you have recurring revenue. It’s whether your product has durable data, regulated workflows, embedded distribution or genuine network effects that an AI-native entrant can’t replicate in a weekend. If that answer isn’t clear, expect a harder conversation with your next buyer than you might have anticipated.

If you run a services or physical-asset business, the conversation has moved in your favour. A founder in 2018 might have struggled to get a meeting for a £10m EBITDA facilities platform. The same founder today is inundated with approacehes. The premium is going to owners who have started to digitise, who have clean data, and who can tell a credible buy-and-build-and-automate story.

A note of caution…

None of this means software is finished. The biggest winners of the AI era will almost certainly be software companies – just not the ones that dominated the last cycle. And the services thesis comes with real risks of its own: labour availability, wage inflation, the complexity of integrating regional operators, and the possibility that AI creates new forms of competition down the line that nobody has fully mapped yet.

But the direction of travel in the mid-market is clear, and we see it every week in the briefs that land on our desks. Private equity isn’t switching off from technology – it’s switching off from the assumption that technology was the only place where durable value could be built.

The next decade’s best deals may well have vans in the car park!