Why the industries hesitating today will have to catch up within two years

In short: 41% of German companies already use AI — in 2024 it was 17%. In some industries the rate is above 70%, in others below 15%. This gap is not a static divide but a widening pair of scissors. For hesitant industries the distance does not shrink while they wait — it grows.

How big is the gap between AI leaders and laggards, really?

The gap is dramatic — and it runs clearly along industry lines. Knowledge-intensive services and research-driven industry lead the way: around 84% of companies in advertising and market research, around 74% in IT services and around 70% in automotive use AI. At the other end sit classic, process-driven industries: construction at roughly 12%, hospitality at around 13%, transport and logistics at about 20% (source: Bitkom / industry analyses 2026).

This is more than a snapshot. Across all companies, AI use has more than doubled in just two years, from 17% (2024) to 41% (2026), with a further 48% planning to adopt it (source: Bitkom AI Study 2026). So the leading industries are not just pulling ahead — they are accelerating.

Why does the gap catch up with hesitant industries — and not the other way around?

Because an AI head start compounds. Whoever adopts AI early gathers data, experience and well-rehearsed workflows — and each of these rounds makes the next application easier and faster. The effect is cumulative: the gap does not grow evenly, it accelerates, because leaders keep converting their lead into new applications. Whoever starts late has to run the same learning curve in less time and under competitive pressure.

For mid-sized companies this means the competitive advantage no longer comes from a single clever tool, but from the ability to build AI systematically into your own value creation. That ability cannot be bought overnight — it grows over time. This is why “adopting later” is not the same as “being just as fast”.

Isn’t hesitation sometimes simply sensible?

It is — and that deserves to be said honestly. Many concerns are valid. A third of companies report that AI was more expensive than expected (Bitkom 2026). According to a Freshworks study, an average of 26% of AI spending in Germany is lost to complexity before any value is created at all (source: Freshworks 2026). And the most common hurdle is not a hype topic but home-made: around three quarters of SMEs struggle with insufficient data quality and data silos, 43% cite data protection and legal uncertainty as an obstacle, and a good quarter cite the skills shortage (source: DIHK digitalisation survey 2026).

Anyone who hesitates in the face of this is not backward, but cautious. The mistake is not the hesitation itself — it is confusing hesitation with standstill. Caution means starting small and in a controlled way. Standstill means not starting at all and hoping the gap will not grow too large. The numbers say the opposite.

What separates real catching up from blind activism?

A clear, tightly scoped first use case — instead of one big AI strategy all at once. Tellingly, only 21% of companies have any formal AI strategy at all, yet 77% of active users report an improved competitive position (Bitkom 2026). Translated, that means value comes from doing and learning, not from the perfect concept in a drawer.

A viable start looks like this in practice:

• Pick a process with real pain — recurring, time-consuming, well-defined. For example quote creation, receipt and invoice pre-checking, first-contact customer inquiries.
• Start small and measurable — one use case, a clear before/after, a timeframe. That way value becomes visible before budget is committed.
• Build in data protection from the start — in the D-A-CH region this is not a brake but a selling point. AI solutions can be built so that sensitive data is anonymized or processed locally.
• Involve the team from the start — onboarding, experimenting and training are part of the use case (more on this in the next section).

This is exactly where weooo comes in: not “pouring AI over something somewhere”, but finding the one process that measurably saves time and cost — GDPR-compliant and built so the solution is actually used day to day. No buzzword bingo, but potential instead of hype.

Why catching up begins with people — not with technology

The technology is rarely the bottleneck — competence in the team is. More than a quarter of companies name a lack of skilled staff as the main hurdle to AI adoption, and many wish above all for more knowledge about sensible fields of use (source: DIHK 2026, IfM Bonn 2026). An AI solution only delivers its value once the people who work with it every day understand it, trust it and know its limits. Catching up is therefore just as much a question of training as of software.

• Experience beats theory: teams build confidence with AI by trying it on real, low-risk tasks — not through a one-off classroom session. A protected space to experiment is worth more than the perfect manual.
• Safe use has to be learned: When can I trust an AI answer, and when do I have to check it? Which data does not belong in a public tool? This judgment is the real AI competence — and it can be built deliberately.
• Competence is no longer a nice-to-have: since February 2025 the EU AI Act (Art. 4) requires employees to have sufficient “AI literacy”. So training your team not only meets a requirement, it lays the foundation for every further use case.
• A culture of trying things out: companies that allow mistakes while experimenting and share what they learn build competence faster than those waiting for the one perfect solution.

This is exactly why, at weooo, onboarding and team training are part of the use case — not an add-on afterwards. A solution nobody can operate confidently saves no time. The most durable advantage does not come from the tool, but from people who use it with confidence.

What does this mean specifically for your industry?

If you work in one of the leading industries, the question is no longer “whether” but “how systematically”. If you work in one of the hesitant industries — construction, logistics, retail, manufacturing, hospitality — now is the cheapest moment: competition there is still thin, and a single well-chosen use case buys a lead that compounds over time. The most expensive path is to catch up under pressure in two years what could be built calmly today.

Conclusion

The gap between AI leaders and laggards is real, it is measurable, and it is growing. Hesitation is understandable — but it does not protect you from catching up, it only makes it more expensive. The companies that cleanly implement a single, well-chosen use case today and bring their team along build a lead that grows with every round. The right first step is smaller than most people think — and right now it is at its cheapest.

Wondering where a sensible first AI use case might sit in your company? Talk to us — we look at your processes and find the point with the greatest leverage.

Sources

• Bitkom, “Artificial Intelligence in Germany” — AI Study 2026 (adoption 41%, strategy 21%, competitive position 77%, costs higher than expected 33%).
• Bitkom / industry analyses 2026 — AI adoption by industry (advertising/market research, IT, automotive vs. construction, hospitality, logistics).
• Freshworks study 2026 — an average of 26% of AI spending is lost to complexity.
• DIHK digitalisation survey 2026 (approx. 5,000 companies) — hurdles: data quality, data protection/legal uncertainty (43%), skills shortage (> 25%).
• IfM Bonn 2026 — knowledge and competence needs in the SME sector.
• EU AI Act, Art. 4 — obligation for employee “AI literacy”, applicable since 02 Feb 2025.