- In 2025, 26 percent of companies in Germany with ten or more employees used AI, and 36 percent of those with 50 to 249 employees (Destatis).
- With a digitalisation strategy, the probability of AI use among German SMEs is 35 percent, without digitalisation activities 19 percent (KfW Research, February 2026).
- The first call takes 20 minutes, is free and runs online.
- Result of the consulting: prioritised use cases, data check, budget in three pots and a 12-month roadmap.
- Since 2 February 2025 the EU AI Act requires measures on staff AI literacy (Article 4).
- BAFA funding for the consulting: coming soon.
What AI consulting delivers for your company
Good AI consulting does not start with technology, it starts with your processes. The result is not a slide deck but a decision: which AI solutions are worth it for your company, in which order, with which budget and under which rules.
In concrete terms you get four things from me:
- Potential analysis: we walk through your business processes and collect use cases from the departments, not from vendor brochures.
- Prioritisation: every candidate is scored on volume, data availability, error tolerance, measurability and effort. That settles what comes first and what can wait.
- AI strategy and roadmap: budgets for pilot, operations and training, responsibilities, approval rules and a twelve-month timeline. What such a roadmap looks like is shown in AI strategy for mid-sized companies in five steps.
- Guiding the rollout: I steer the first AI pilot project, measure it against targets agreed in advance and decide with you after 90 days whether to expand or stop.
Tools come into play once it is clear which problem they should solve. That is why I choose them to fit your process, not the other way round.
AI in mid-sized companies: where businesses stand today
Artificial intelligence has arrived in German mid-sized companies, but unevenly. According to the ICT survey of the Federal Statistical Office (Destatis), just over a quarter of companies with ten or more employees used AI in 2025, and almost six in ten of the large ones.
| Company size | Share using AI in 2025 |
|---|---|
| 10 to 49 employees | 23% |
| 50 to 249 employees | 36% |
| more than 250 employees | 57% |
| all companies with 10 or more employees | 26% |
Source: Destatis, ICT survey of enterprises 2025
The gap has a reason, and it rarely lies in the budget. In February 2026 KfW Research analysed what drives AI use among German SMEs: companies with a digitalisation strategy use AI with a probability of 35 percent, companies without any digitalisation activities with 19 percent. Company size plays only a minor role. The study names a lack of skilled staff and competence, limited time, insufficient data and concerns about security and reliability as the main obstacles.
This is exactly where AI consulting helps. It does not replace skilled staff, but it saves the months in which a team tries out tools without a plan.
How AI consulting works
The consulting is split into four steps. After each step you decide whether to continue.
| Step | What happens | Result |
|---|---|---|
| 1. First call, 20 minutes, free | You tell me where things get stuck. I tell you whether and how AI helps. | Clarity on whether consulting is worth it |
| 2. Potential analysis in an AI workshop | Map processes, collect use cases, check data, on site or online | Scored list of candidates |
| 3. AI strategy and roadmap | Prioritise, set budget, responsibilities, data protection and the AI Act | Twelve-month roadmap ready for a decision |
| 4. Guiding the rollout | Steer the AI pilot project, measure targets, bring the team along | Decision on expansion or stop after 90 days |
How long this takes depends on size and data. Between the first call and a finished roadmap there are usually a few weeks. I give you a binding estimate of the effort after the first call so you can plan.
AI use cases for mid-sized companies that pay off first
The first AI projects should be small, frequent and checkable. From my own practice, these work particularly well:
- Incoming invoices and receipts: extract data, check against the order, flag deviations.
- Customer service: draft replies to recurring requests that a person approves. What is realistic there is covered in What an AI agent can really do in customer service.
- Quotes and tenders: monitor public tenders, prepare quotes from templates.
- Company knowledge: make manuals, contracts and minutes searchable, if you like with AI on your own servers.
- Reporting and sales: weekly reports, market monitoring and data hygiene in your CRM.
Whether a simple AI assistant is enough or an AI agent that carries out several steps on its own makes sense is decided case by case. I explain the difference in What an AI agent really is.
AI consulting and implementation as a team
I deliberately keep consulting and implementation separate and closely linked. I advise on strategy, selection and rollout. If automations or AI agents are to be built afterwards, the team at Weidner & Friends takes over the implementation. Founders in their first years can turn to the business consulting at Gründerfreunde (in German).
Your advantage: one contact from the first question to live operations, and a roadmap that holds just as well if you implement it with your own IT or another partner.
Data protection, the AI Act and AI literacy in AI consulting
Every AI rollout needs a framework, and it is smaller than many fear. Three points belong in every consulting project:
- Data protection: which data may leave a system, where it is processed and whether a data processing agreement under Article 28 GDPR is in place.
- AI Act: since 2 February 2025, Article 4 of the EU AI Act requires measures on staff AI literacy. High-risk uses, such as pre-screening job applications, face additional obligations from 2 December 2027.
- Co-determination: where a works council exists in Germany, it is informed during planning (section 90 BetrVG) and has a say on systems that can evaluate behaviour or performance (section 87(1) no. 6 BetrVG).
An AI inventory, clear responsibilities and fixed approval rules usually cover this for a mid-sized company. Where legal advice is needed, I tell you.
BAFA funding for AI consulting: coming soon
Small and medium-sized companies in Germany can have business consulting subsidised by the Federal Office for Economic Affairs and Export Control (BAFA), under the programme “Förderung von Unternehmensberatungen für KMU”. It funds exactly what good AI consulting delivers: analysis, concrete recommendations and guidance on implementation.
BAFA funding for my AI consulting is coming soon. Mention it in the first call and we will plan the consulting from the start to fit the programme.
Why work with me as your AI consultant
I have been building companies for more than 20 years, from my time at Rocket Internet through CryptoTicker to my holding Weidner Ventures. In my own companies AI agents work alongside the team every day, in research, editorial work, reporting and sales. Before a process gets a new hire, we check whether an agent can reliably take it over. So whatever I recommend, I have tried myself, mistakes included.
More about how I work is on the page AI consultant Dennis Weidner.
What does AI consulting cost?
The first call is free. For every further step you get a fixed price before we start, depending on size, number of departments and data. The cost of the AI systems themselves only arises during implementation; the largest item there is rarely the model but the integration with your systems and ongoing operations. With BAFA funding, which is coming soon, a grant covers part of the consulting costs.
Frequently asked questions
What does AI consulting do?
It finds the processes where artificial intelligence saves measurable time or money, checks data, costs and the legal framework, and sets a roadmap. The result is prioritised use cases, a budget and clear responsibilities instead of a list of tools.
Which companies is AI consulting suitable for?
For small and medium-sized companies that want to introduce AI with a plan rather than on the side. Under the EU definition these are companies with fewer than 250 employees and annual turnover of no more than 50 million euros or a balance sheet total of no more than 43 million euros. You do not need your own IT department.
What does AI consulting cost?
The 20-minute first call is free. For the following steps I give you a fixed price in advance, depending on size and data. BAFA funding for the consulting is coming soon.
How long does an AI consulting project take?
Usually a few weeks from the first call to a finished roadmap. A first AI pilot project then runs for about three months before the decision to expand or stop.
How do AI consulting and AI training differ?
Consulting decides where and how AI is used in the company. Training teaches staff to work with it safely and supports AI literacy under Article 4 of the AI Act. Both belong together, so I plan training into the roadmap.
Will AI replace management consultants?
Parts of the work, yes: AI now handles research, market overviews and first drafts in a fraction of the time. What remains is the decision and the responsibility for it. That is exactly why you need a consultant who uses AI himself.
Further reading
- AI strategy for mid-sized companies: a five-step roadmap
- AI consultant Dennis Weidner
- Local AI for SMEs
- My AI stack as a founder
- AI in the founder's working day
Note: AI tools supported me in writing this page. I stand behind its content and every statement with my name.