A good platform without use is worthless. The return does not appear when the
licence is bought but on the day your employees open the AI voluntarily. That is
not a technical question but a human one. In our AI adoption work, we see the
same pattern: adoption decides the ROI — not the choice of model, not the feature
set.
86 %
AI usage at LAUDA
76 %
use it daily
~1,000 h
saved / month
7.5
FTE equivalent
LAUDA · manufacturing technology · public case study
In most projects the discussion circles around "GPT or Claude?". The actual
increase in value comes from the usage rate, though, not from the last three per
cent of model quality. The distance between a weak and a strong rollout is not
"ten per cent more output" — it is a factor of ten. The best model that nobody
uses returns nothing.
You don't reach 76 % daily usage through mandatory training, but through real
value in everyday work.
Three to five enthusiastic employees drive adoption across the whole company. No
budget in the world replaces them. At LAUDA these people are called AI
ambassadors — one key user per department. The expectation there is concrete:
five to ten new workflows per department per month. That creates momentum which
sustains itself.
A fixed rhythm belongs to it. LAUDA runs a two-week release cycle with direct
user feedback. Small, visible results every fortnight keep the initiative alive —
and the champions on board.
Do not set unrealistic expectations. At LAUDA the path from scepticism to
enthusiasm took seven to eight months. That is not a failure but the normal
case. Expecting full acceptance after six weeks wrongly declares a working
rollout a failure.
Your employees have been using AI for a long time — just uncontrolled, through
private accounts. A ban without an alternative does not solve that, it sharpens
it. The right order is: the legal platform first, then the rule for private use.
If the internal solution is at least as good as ChatGPT, people move back to it
on their own.
Banning shadow AI without offering an alternative makes the problem worse.
A tool without support — an industrial company of around 800 employees
licensed an AI assistant for everyone: one all-staff email, one FAQ, a
20-minute slot at the town hall. After 14 months usage stood at 7 %, the best
use case was "politer emails", and the return was negative.
Consulting without a platform — a service provider bought a 180-page
strategy paper. After 18 months: two half-finished prototypes, no rollout. A
roadmap without a shared tool is a pile of sticky notes.
The endless pilot — without a clear go/no-go gate the pilot becomes a state
rather than a step. The champions tire, the sceptics are vindicated.
No champion — the business case was sound, the budget was free. What was
missing was the one person who carries the topic internally. Without them every
initiative dies.
The gap in the market is large: according to Bitkom 2025, 43 % of companies
offer no AI training, and 70 % of people in work have never received any
development. This is exactly where your adoption is decided. Count training as an
adoption multiplier, not as a line in a cost centre. The consequence for you:
write down your three to five champions before you buy the platform. If no name
comes to mind, start there — not with the tool.
Thanks to meinGPT, introducing artificial intelligence went smoothly.
Change management is the human side of the rollout. How to build the champions
structurally and anchor learning directly in the tool is shown by
enablement & training. How it all fits the overall plan is in the
strategy. And what the usage is finally worth is calculated by
measuring impact.
ISO Certified
GDPR Compliant
EU Hosting
Start with AI in your company
Together we find the right use cases, connect your systems, and bring AI into daily work in line with your business.