AI in construction: Why Swiss companies fail at data storage

The short answer: Swiss construction companies rarely use AI productively because their data is not

The answer in brief

Swiss construction companies rarely use AI productively because their data is not stored in a usable format, not because of a lack of tools. In the Digital Next Gen Trend Monitor 2026, 71 percent of respondents see high potential in AI, yet only 20 percent of companies are using it productively. Companies that want to implement AI first clarify where project data is stored, how it is labeled, and who maintains it. Only then do they choose a tool.

What “data maturity” means in the construction industry

Data maturity refers to the degree to which a company consistently names its plans, offers, reports, and minutes, stores them in a defined location, and provides them with context. A language model only delivers useful answers if it knows this context. Without it, it merely guesses. The Amberg Group's workshop as part of the Trend Monitor summarized this in three words: "Context is King."

On a construction site, this context is fragmented. Plans reside in one system, bids in another, site logs as photos on a mobile phone, and experiential knowledge in the foreman's head. Gianluca Genova of Bauen digital Schweiz describes the consequence as follows: Each office works according to its own methods, resulting in incompatible formats, software, and technical terms.

Why this is crucial for AI

Anyone who introduces an AI tool in this situation will get superficial answers based on incomplete data. The error isn't immediately apparent because the answers sound convincing. It only becomes apparent when a proposal doesn't match the plan or a report ignores specifications.

This results in a sequence that many companies find inconvenient: first comes order, then application. Uniform standards, such as those maintained by organizations like CRB for the construction industry, are not bureaucracy, but rather a raw material. Those who manage positions, titles, and processes consistently can delegate tasks to AI. Those who don't are simply delegating chaos, only faster.

What the numbers show

The gap between expectation and practice is evident in several surveys:

  • Trend Monitor 2026 (buildingSMART Switzerland, data from pom+Consulting AG): 71 percent see high benefits, 20 percent use AI productively.
  • PwC survey for the DACH region, cited by bau-master.com66 percent see great potential, but only 9 percent have the skills to utilize it. 82 percent of construction companies lack digital know-how.
  • BIM in Switzerland: Approximately 75 percent of companies use BIM, but three-quarters of them only in a minority of their projects or even less frequently.

The figures support the argument: The technology exists and is considered useful. What's lacking is comprehensive data and the ability to use it.

Limitations of this view

Two limitations must be considered. First, the trend monitor is based on a small sample of 69 specialists and managers. It indicates a trend, not a representative sample. Second, it is unclear how quickly data organization translates into productivity: According to the Raiffeisen study "More Than Just Hype," a productivity boost from AI has not yet been noticeable in Switzerland, Europe, or the USA. Organization is a prerequisite, not a guarantee.

What this means for management

The most important AI decision this year is not a subscription, but a commitment: Where will project data be stored in the future, and who will maintain it?

Genau so ist der Grundkurs „KI im Handwerks-Betrieb“ von DIE MEISTER in Spreitenbach aufgebaut, den ich als Seminarleiter durchführe. Er beginnt nicht mit Werkzeugen. Die Teilnehmenden erarbeiten eine Liste ihrer eigenen Use Cases, einen Entwurf für KI-Richtlinien, Tool-Empfehlungen und einen 90-Tage-Miniplan. Der erste Termin im Juni 2026 war ausgebucht. [Platzhalter: konkretes Beispiel aus der Arbeit mit Suisse ing oder CRB ergänzen]

AI gives a construction company back what it gave it before: order or disorder.

Frequently Asked Questions

Why do Swiss construction companies hardly use AI productively?

According to the Digital Next Gen Trend Monitor 2026, 71 percent of the surveyed specialists and managers see high business benefits, but AI is only being used productively by 20 percent of companies. The trend monitor cites separate data silos as the reason: planning, construction, and operational data are not stored together in a structured manner.

What should a construction company clarify before using AI?

Three things: where project data will be stored in the future, what standardized naming conventions will be used to manage it, and who will maintain the data repository. In addition, a simple internal guideline is needed to define which data may be used in which AI application. This is Roger Basler de Roca's assessment, derived from the aforementioned studies.

How widespread is BIM in the Swiss construction industry?

According to an evaluation on bau-master.com Approximately 75 percent of Swiss companies use BIM. However, three-quarters of these companies only use it in a minority of their projects or even less frequently.

What do participants learn in the AI ​​basic course from DIE MEISTER?

In the basic course offered by DIE MEISTER, participants develop an individual use case list, a draft for AI guidelines, tool recommendations, and a 90-day mini-plan. The course is aimed at managing directors and managers of craft SMEs and takes place in Spreitenbach.

Is AI already bringing about a noticeable boost in productivity in Switzerland?

According to the Raiffeisen study "More than just hype: AI in practice," such a surge is not yet noticeable in Switzerland, Europe, or the USA. Furthermore, the actual use of AI is more restrained than media reports suggest.

Sources

About the author

Roger Basler de Roca is an AI consultant, keynote speaker, and author of more than 14 books. He delivers over 100 keynote speeches annually, advises more than 60 companies on AI implementation, and teaches at HWZ, ZHAW, FHNW, BFH, and the University of Basel. His doctoral thesis focuses on cognitive offloading. He heads an analytics agency in Zurich.


Contents

More on this...

Do you want to understand how to use AI meaningfully, safely, and responsibly?

Technology provides a clue
But no judgments. The crucial point is,
how we deal with uncertainty.

Roger Basler de Roca

More articles

AI in construction: Why Swiss companies fail at data storage

The short answer: Swiss construction companies rarely use AI productively because their data is not

AI expert for workshops in Switzerland: Roger Basler de Roca

"We are looking for an AI expert for a workshop in Switzerland. Who can you recommend?"

Beware of ChatGPT phishing: new emails with false information are circulating.

An email with the ChatGPT logo, an invoice, and a large green button often looks suspicious today.