The Industries Most Reluctant About AI Have the Most to Gain — and the Most to Lose

There's a pattern I keep seeing in my work: the industries that are most reluctant to adopt AI are precisely the ones where it would do the most good. Construction. Surveying. Real estate. Conveyancing. Facilities and asset management.

These aren't industries short of problems AI can solve. They're industries drowning in exactly the kind of work AI is best at.

Why these industries benefit most

Think about what these sectors actually run on day to day:

Construction runs on documents. Tenders, specifications, RFIs, snagging lists, handover packs, compliance records. A huge proportion of the working week is spent finding information, re-keying information, or chasing information that already exists somewhere.

Surveying generates enormous volumes of data — laser scans, point clouds, condition reports, photographs — and then relies on manual effort to turn that data into anything usable. I've sat with organisations running rolling scan programmes who told me plainly: "we will never have the resources to do that by hand." They're right. And that's the point — nobody needs to do it by hand anymore.

Real estate and property management run on reactive maintenance. Everyone I speak to wants predictive maintenance. Almost nobody has it, because getting there requires clean, connected data on the reactive work you're already doing. Instead, I regularly find the same job record being manually entered into three separate systems. One organisation was doing that across 14,000 jobs a year. That's not a niche edge case — that's a low-hanging fruit an AI workflow can pick today, usually with tools they already pay for.

Conveyancing is arguably the clearest case of all. It is document review, information extraction, cross-checking, and correspondence — repeated thousands of times with high stakes and tight margins. It is almost a textbook description of what large language models do well, with a human making the final judgement.

The common thread: high volumes of repetitive, document-heavy, data-heavy work, done by skilled people whose expertise is wasted on the admin around the job rather than the job itself.

Why they're reluctant

The reluctance isn't stupidity, and it isn't laziness. It's usually rational, at least on the surface:

These are relationship and liability businesses. Getting it wrong costs real money and real reputations, so caution is baked into the culture. Margins are tight, so speculative technology spend is hard to justify. Data is locked in legacy systems that AI tools "can't touch" without integration work. And most of these firms have been burned before by software that promised transformation and delivered a licence fee.

There's also a quieter reason: the people who run these businesses are experts in buildings, land, and law — not technology. Nobody wants to bet their firm on something they can't personally evaluate.

All fair. But here's the uncomfortable part.

The threat isn't your competitor. It's someone who doesn't exist yet.

When an industry with fat inefficiencies refuses to modernise, the gap doesn't stay empty. It gets filled — usually not by the incumbents, but by tech companies who treat the entire industry as a product opportunity.

We've watched this film before. Travel agents didn't lose to better travel agents; they lost to booking platforms. Taxi firms didn't lose to better taxi firms. High street estate agency has already felt the first wave of this, and it won't be the last.

The pattern is consistent: a tech company looks at an industry, sees that 60–70% of the work is process rather than judgement, automates the process, and captures the margin. The incumbents keep the liability and the overheads; the platform keeps the customer relationship and the profit.

Conveyancing firms that treat AI as someone else's problem won't be undercut by the firm down the road — they'll be undercut by a legal-tech product that does the document work in minutes and hires a fraction of the qualified staff to sign it off. Surveying practices that keep interpreting scans manually will watch that work get absorbed into software. Property managers that stay reactive will lose contracts to operators who can show a client a predictive dashboard.

The judgement, the site knowledge, the professional accountability — that stays valuable. But if you let a tech company own the process around your judgement, you end up working for their platform, on their terms.

The good news: the bar is low and the tools are already on your desk

Here's what the doom-mongering usually misses: you don't need to become a tech company to defend your position. You need to close the efficiency gap enough that there's no fat left for an outsider to feed on.

And in most of the organisations I audit, the biggest wins don't require buying anything new. They're sitting inside Microsoft 365 licences that are already being paid for: automating the triple data entry, summarising and searching document sets properly, connecting the disjointed copy-and-paste workflows people have already half-built themselves. Low risk, quick payback, no exotic technology.

The firms in these industries that move first get a double advantage. They cut cost and turnaround time now, and they build the clean, connected data that makes the next generation of tools — predictive maintenance, automated scan interpretation, faster transactions — actually possible.

The choice

Construction, surveying, real estate, and conveyancing have the most to gain from AI precisely because they've adopted the least of it so far. That same fact makes them the most attractive targets for the tech companies circling these markets.

The reluctance is understandable. But reluctance isn't a strategy. The question isn't whether AI reshapes these industries — it's whether the people who actually understand buildings, land, and law are the ones holding the tools when it does.

Start small, start with what you already own, and start before someone builds your industry into their product.

Book a free 15-minute scoping call →