Why AI in construction is only now starting to land, when the technology has existed for years

The short answer. Construction has been known since at least 2017 as one of the least digitised sectors in the world, and productivity in construction grew by only 10% worldwide between 2000 and 2022, against about 50% for the total economy (McKinsey, 2024). Still, things are moving: investors put around 50 billion dollars into construction technology worldwide between 2020 and 2022, 85% more than in the three years before (McKinsey, 2023). In the Netherlands, AI use in construction lags behind: 8.9% of companies in the Dutch construction industry used AI in 2024, against 22.7% of all Dutch companies (Statistics Netherlands, CBS). So the technology has existed for a long time. Our explanation: AI is only starting to land now that it is applied to the work itself, such as estimating and work preparation, instead of only alongside it.

How long has construction been lagging behind?

The best-known research on this is Reinventing Construction: A Route to Higher Productivity by the McKinsey Global Institute from 2017. It describes construction as one of the least digitised sectors in the world: next-to-last of all sectors in the United States, and last in Europe. Labour productivity in construction grew by an average of 1% a year over two decades, against 2.8% for the world economy and 3.6% for manufacturing. Less than a quarter of construction firms kept pace with the productivity growth of their own national economy.

A 2024 follow-up by McKinsey shows this was no snapshot. Between 2000 and 2022, productivity in construction grew by only 10% worldwide, against about 50% for the total economy. In the United States, construction productivity even fell by an average of 2% a year between 2000 and 2021, while in China it grew by 4% a year.

Why is construction lagging behind?

A few causes come up in almost every study:

Historically low IT investment. Construction firms have historically spent less than 1% of their revenue on IT, according to McKinsey.

A fragmented market. Dutch construction has an above-average number of small companies, while it is the large companies that do more with digitalisation, writes the Dutch bank ABN AMRO in 2024.

Knowledge and culture. The Dutch construction employers’ association Bouwend Nederland identified culture change, high investment costs and a lack of digital knowledge within the company as the biggest obstacles. ABN AMRO adopts those findings.

Time pressure and tight budgets. Working with tight budgets and under time pressure leaves little room for a culture in which people dare to experiment. According to an earlier ABN AMRO survey in the Netherlands, time pressure is also the most frequently cited cause of failure costs in construction.

What is changing now

There are concrete signs of movement:

• Investors put around 50 billion dollars into construction technology (AEC tech) worldwide between 2020 and 2022, 85% more than in the three years before, according to McKinsey’s From start-up to scale-up (2023).
• In the Netherlands, the share of construction site workers who use a digital system daily grew from 43% in 2018 to 53% in 2022. That comes from the Netherlands Working Conditions Survey by the Dutch research institute TNO, as analysed by ABN AMRO.
• In the Netherlands, 33% of construction firms use digital systems for Building Information Modelling (BIM), according to research by Bouwend Nederland cited by ABN AMRO.

At the same time, construction lags behind on AI. Across the Netherlands, 22.7% of companies with ten or more persons employed used some form of AI in 2024, against around 14% in 2023. In the Dutch construction industry, the figure was 8.9% in 2024, one of the lowest of all industries (CBS, 2025). So construction benefits less from the acceleration in AI than the rest of the economy, even though the sector is firmly in the sights of investors worldwide. More figures on AI in Dutch business are in the state of AI in Dutch SMBs.

Why it has not really landed yet

The productivity figures show that the investment has not yet led to a clear leap in productivity. McKinsey points to one explanation in 2024: technology in construction has mainly been used for control, and less for productivity.

Our reading of that, and this is our interpretation, not McKinsey’s: many tools track the work, with progress tracking, dashboards and reports, but do not take over the underlying work. Think of estimating, work preparation and the handover between systems. That is where the manual work is, and where a lot of failure costs arise.

That is the gap between “the technology exists” and “it works on the shop floor”. Money and attention are there now, but most tools watch instead of working along. That is why AI in construction is only now starting to land: not because the technology is new, but because it is only now being applied where the real manual work is. That does require something of the foundations: project data that is correct in one place, with a data foundation, and systems that talk to each other, by connecting your systems. Also read why so many AI projects deliver nothing.

What this means in concrete terms for contractors, installers and subcontractors is covered on AI in construction.

Sources

• McKinsey Global Institute, Reinventing Construction: A Route to Higher Productivity (2017)
• McKinsey, From start-up to scale-up: Accelerating growth in construction technology (2023)
• McKinsey, Delivering on construction productivity is no longer optional (2024)
• ABN AMRO, Digitalisering in bouw neemt toe, maar nog veel te winnen (April 2024, in Dutch)
• CBS, Gebruik kunstmatige intelligentie (AI) door bedrijven neemt toe (February 2025, in Dutch)

Frequently asked questions

Why is the construction sector lagging behind on digitalisation?

Mainly because of historically low IT investment (less than 1% of revenue), a fragmented market with many small companies, a lack of digital knowledge, and tight margins and time pressure that leave little room to experiment.

How many construction companies in the Netherlands use AI?

According to Statistics Netherlands (CBS), 8.9% of companies in the Dutch construction industry with ten or more persons employed used some form of AI in 2024. Across all industries, it was 22.7%. That puts construction among the sectors with the lowest AI use.

Is this changing now?

Yes. Worldwide, around 50 billion dollars went into construction technology between 2020 and 2022, 85% more than in the three years before (McKinsey, 2023). In the Netherlands, the share of construction site workers who use a digital system daily grew from 43% in 2018 to 53% in 2022.

Why has that investment not led to more productivity yet?

According to McKinsey (2024), technology in construction has mainly been used for control, and less for productivity. Our interpretation: most tools track the work, but do not yet take over the underlying work, such as estimating, work preparation and the handover between systems.