When the groundwork is done — and the numbers still don't show it
- Client
- Regnauer Fertigbau GmbH & Co. KG, Seebruck/Chiemsee (DE)
- Industry
- Prefabricated house construction / manufacturing
- Period
- August 2026
Starting point
Regnauer Fertigbau moved early and consistently on AI: tools made available to everyone, a dedicated AI role staffed, a weekly format for passing on knowledge, plus tutorials and a podcast. In sales and architecture, AI is already in productive use — visualisations and material variants that used to be budgeted in hours now take minutes.
Only the usage figures lagged behind that groundwork. Management wanted to know why, and commissioned a workshop to show what is possible and to encourage use.
Approach
Instead of training straight away, we listened first. In a single day: four one-to-one conversations with managers from sales, production, engineering and the AI role, 30 minutes each, deliberately set up as a diagnosis rather than a product demonstration — plus a closing conversation with management.
Two documents came out of it: an AI use case map with around 30 applications across all areas, assessed by benefit, effort and prerequisites, and a report to management with prioritised recommendations across three time horizons and six concrete decision questions.
Result
No acceptance problem — quite the opposite. Every manager interviewed was already working confidently with AI, and each of them brought concrete, economically sensible ideas for applications. The tools and measures in the company had been chosen well. What kept the curve flat were two other things:
Too much choice
As long as it is not laid down which tool is meant for what and which content may go into it, choice does not lead to a decision but to hesitation. A one-page classification solves that — a decision, not a project.
The format, not the content
The weekly short format reaches around sixty people and is strong in content, but generates hardly any questions: in front of a group that size, nobody likes to ask. That is exactly why the one-to-one conversations worked — an incidental finding the client adopted straight away. Change management moves faster in a personal format than in a large round.
In addition: a named point of contact for ideas, with a small budget for trying things out. Incidentally, the approach also satisfies the obligation to demonstrate AI literacy under Article 4 of the AI Act.
The leadership workshop confirmed that we at Regnauer are already working sensibly with AI. At the same time we identified clear priorities and concrete next steps in order to direct usage more purposefully and to make use of its potential for real efficiency gains.
What we take from it
Flat usage figures are rarely a motivation problem. More often they are a sign that good groundwork is not yet focused: too many options without assignment, and knowledge shared in too large a round.