AI in construction: why the website still comes first

In July 2026, ANCE, Italy's national association of construction companies, gave its member companies the first AI tools for reading bills of quantities, assessing tenders and checking site safety documentation. That's a real step, not an empty announcement. But the latest official data tells a different story about the sector as a whole: construction is one of the industries furthest behind on AI adoption, and roughly one construction company in three, among those with at least 10 employees, still has no website of its own. Two facts describing the same gap from different angles: a trade association building advanced tools for its members, while part of that same membership is missing the basics.

The tools ANCE has made available

The solutions released in July 2026 come out of the AI-ANCE working groups, built with input from more than forty member companies. They cover three areas: a module that lets you upload a bill of quantities as a PDF and query its contents in natural language, a beta tool for a preliminary assessment of public tenders, and a system that automatically checks the completeness of site safety documentation, from risk assessment documents to mandatory hiring notices and training certificates.

These are tools reserved to member companies, built for internal work, the kind done with contracting authorities, suppliers and job sites. They have nothing to do with communicating with customers, and ANCE hasn't yet published any figures on how many companies are using them or with what effect: the rollout is only a few weeks old, and it will take time before a clearer picture emerges.

What the ISTAT data says about the sector

The ISTAT report "Imprese e ICT 2025" measures, for the first time at this level of detail, how far AI has actually spread inside Italian companies with at least 10 employees. The national picture is fast growth: from 5.0% in 2023 to 8.2% in 2024, up to 16.4% in 2025. In construction, though, the share stops at 10.4%, the second-lowest of the eleven sectors ISTAT measured, ahead only of water supply and waste management. Transport and warehousing, a neighbouring sector in terms of the kind of work involved, sits at 11.6%.

ISTAT also points to a structural cause that isn't specific to construction: nearly 60% of companies that considered but didn't go ahead with an AI investment name a lack of adequate skills as the main reason. It isn't a judgment on these businesses, it's a snapshot of a transition that, in some sectors, moves more slowly because it needs skills that are currently missing, not because interest is missing.

The gap isn't only about AI

The same ISTAT report also measures something simpler: how many companies have a website of their own. In construction, the figure is 65.3%, against a 76.5% average across all sectors, meaning roughly one construction company in three, in that size bracket, still has no online presence of its own. That's a smaller gap than the one on AI in percentage terms, but it concerns something far more basic: not advanced automation, but the simple possibility of being found by someone searching for the company by name.

That holds even for companies working mostly on public tenders rather than with private customers: a website helps subcontractors and suppliers find the company, shows completed work to whoever is assessing it for a tender, and makes it reachable to people looking for work. I've written about a related case, artisan businesses that work on commission, in website for artisans: the type of customer changes, the logic doesn't.

Why the website comes before the AI

An AI tool, however useful, works inside processes a company already has: it speeds up reading a bill of quantities, helps assess a tender, checks documents that already exist. It doesn't replace the basic fact of being reachable online, which is a different, more upstream problem. A company investing in internal automation before having a way to be found is building the second floor before the first: it works, but not for whoever comes looking from the street.

It isn't a question of which one costs more: I've written the market price ranges for a business website in Italy in full in how much does a website cost. It's a question of order: the basics first, then whatever gets built on top of them.

Where to start

For a construction company with a fixed office or yard, being found starts with two simple things: a website that shows completed work, certifications and areas of operation, and a Google Business Profile for people searching for a company nearby. I've written about this in detail in ranking on Google Maps: the factors Google says it uses, and what actually makes a difference for a local business. Only once the basics are in place does it make sense to look at what to automate in day-to-day work.

If you want to work out where your company actually stands and what's worth fixing first, get in touch: tell me how you work today, with whom, and I'll tell you what's worth sorting out first.