AI strategy for SMEs: From initial understanding to concrete implementation

What is AI actually — beyond the buzzword?

In essence, it is about AI systems taking on tasks that would normally require human intelligence, such as recognizing patterns, preparing decisions, understanding language, or making predictions.

Three AI technologies that are relevant for SMEs:

Rule-based AI works according to clear “if-then” rules. Perfect for standardized processes such as the automatic sorting of inquiries or the review of documents.

Machine learning means: Systems learn from data and improve over time. This technology helps identify patterns in business data — from customer behavior to machine maintenance planning.

Generative AI is currently the most visible form (ChatGPT & Co.). She creates content, answers questions, and helps with content creation. Particularly interesting: Almost three quarters of SMEs using AI are already using generative AI.

The good news: You don't have to understand everything at the same time. The use of AI in SMEs does not mean becoming a tech company overnight. It means strategically identifying the areas where AI tools can help. Decision-makers also regularly report on these experiences in our d:u podcast.

‍

Where AI can help in practice — three key questions

From our discussions with SMEs, we know that there are areas almost everywhere where AI can create real added value. The question is not if, but where.

1. Where is time regularly lost with recurring tasks?

In almost every company, there are jobs where teams go through the same steps over and over again; sorting requests, transferring information, or creating similar documents.

This is where AI unleashes its greatest potential. Automating recurring processes, speeding up processes and streamlining complex structures saves considerable time. Teams can invest this gained time in strategic and value-adding tasks.

2. Is there any data that has untapped potential?

Most SMEs collect data, such as customer data, sales figures, process information or machine data. Only a few actually use this database. Efficient use of existing data is a key opportunity here.

This data often contains valuable information: Which customers will migrate? Which products will sell well? Where do bottlenecks arise? AI models can help make these patterns visible.

3. Where are important decisions made on an uncertain basis?

Whether in pricing, purchasing, personnel planning or product development, data-based decision support can make a real difference here. AI provides well-founded foundations where, up to now, decisions have often been based on gut feeling.

‍

Where AI offers great potential

The successful use of AI depends on the right context. AI systems develop their full potential particularly when:

• Processes are repeated frequently and run according to recognizable patterns - the more standardized, the better
• Digital data is available, which are currently not fully exploited
• loss of time or error Really cost money in certain areas
• Teams tied up through routine work are and need more capacity for strategic activities

In many medium-sized companies, several of these points apply. The question is not whether AI can help, but where is the best place to start. An exchange with other SMEs is extremely valuable at this point. AAt the data:unplugged festival, we make it possible to exchange ideas with other companies about concrete practical examples.

‍

How to find the right starting points in the company

Instead of reacting to external trends, companies should systematically analyze where AI provides the greatest benefit. A structured approach helps to:

Step 1: The internal inventory

Have conversations with the team. Not about AI, but about daily work. What costs unnecessary time? Where is there frustration? Where do errors occur?

These discussions often reveal surprisingly clear starting points. From our experience: The best AI projects are created where teams themselves see the biggest pain point.

Step 2: Prioritize by impact

Of all the challenges that arise, which would have the biggest business impact if solved? This is where the lever lies.

Also take into account the existing AI skills in your teams. Four out of five companies lack basic AI skills. This is normal and not an exclusion criterion.

Step 3: Identify quick wins

Are there processes that would be relatively easy to automate and show results quickly? These are the perfect way to get started. You build up experience and thus build trust in the team.

Examples from practice: AI-supported marketing emails, an AI chatbot for standard customer inquiries, or tools for analyzing customer data.

Step 4: Specify options

For the most promising starting point: Which AI applications already exist? What have others in the industry done? What would implementation mean in practice?

The company should ensure that employees have sufficient AI skills, as certain AI practices are prohibited. Check at an early stage what requirements apply to your planned AI systems in order to comply with legal and ethical requirements.

Step 5: Start small, learn quickly

Start with a manageable pilot project and set clear criteria for success. Stay open to readjusting and optimizing.

Dealing with new technology requires patience and a willingness to learn. This approach helps to find relevant starting points.

‍

What AI tools can actually do

More specifically, AI won't transform your business overnight, but it can bring real, measurable improvements in many areas.

Save time on routine tasks

AI software can significantly reduce repetitive work. The automatic pre-sorting of inquiries, the pre-qualification of applications, the categorization of documents — all of this often costs teams time that could be used for more demanding tasks.

Better basis for decision-making

Instead of relying on gut feeling, decisions can be made based on data analyses. Which marketing channels really produce results? Which customers have the highest potential?

Companies in particular hope for faster data analyses and more innovation through the use of AI - Around 70 percent mention these aspects in each case.

Error reduction and faster response times

Where people might get tired and make mistakes when doing monotonous tasks, AI systems work constantly. This can significantly reduce the error rate during audit, data entry or quality control.

In customer service, AI-powered chatbots can reduce waiting times and increase satisfaction. Integrating such solutions is often easier than expected.

Scaling without proportionally increasing effort

As businesses grow, personnel costs usually also grow proportionally. AI technologies can help scale specific areas without the need to hire linearly more people.

This applies to marketing, sales, customer service, and many other areas. SMEs expect a return on AI investments of a factor of four within one year.

These effects are not theoretical scenarios, but can be realized with suitable AI systems. Practice-oriented approach and realistic expectations are decisive.

‍

The first steps to implementing AI

No one has to go from zero to a hundred. A strategic, gradual start is not only smarter — it is also more successful. Here's a realistic roadmap:

Stage 1: Understanding (1-2 months)

Continuing education is important. Not through theoretical courses, but through discussions with other SMEs who are already using AI. At the data:unplugged festival, we specifically promote this exchange between decision makers and data experts. How do they actually implement AI? What legal hurdles have they overcome? Where have they perhaps even failed?

You should also understand the legal framework. Teams should develop a basic understanding of the role of AI.

Stage 2: Analyze (1 month)

Collect specific examples from your own company. Document how much time certain processes take. Calculate what an improvement of 20%, 40%, or 60% would mean.

Based on this basis, identify where the integration of AI software would have the greatest benefit — whether in marketing, production, sales, or maintenance and repair.

Phase 3: Probe solutions (1-2 months)

Once a specific starting point has emerged, take a look at 2-3 different AI platforms or tools. Schedule demos and check references.

Calculate ROI potentials, taking into account key factors such as data protection requirements, the quality of the existing database and the ability to integrate into existing systems.

Phase 4: Piloting (3-6 months)

Start with limited use in a specific area. Gain experience, measure the results and adjust if necessary.

Pay attention to how it works in everyday work and get feedback from the teams. How to use the new technology must be learned, both technically and culturally.

Only when the pilot is successful should you scale up to other areas of application. After six months at the latest, you will not only have theoretical knowledge, but also concrete experience and measurable results.

‍

What is possible with AI tools: practical examples

For SMEs, the use of artificial intelligence is worthwhile when there are clear goals and it is known where to start.

In customer service Service teams can obtain more capacity for complex cases through automated pre-sorting with AI chatbots. A third of German SMEs are already using AI, many of them in the service sector.

In production Predictive maintenance can reduce unplanned outages and reduce maintenance costs. Quality problems can be identified earlier and counteracted through intelligent analyses.

In administration Routine processes can be automated and thus capacity can be created for growth-relevant tasks. Repetitive tasks in accounting can be reduced through AI-based automation.

In sales It is possible to predict more precisely which leads have potential. Marketing can be invested more specifically, as data-based analyses show which measures work — from creating images to personalized marketing emails.

The various AI applications open up a wide range of options. The opportunities are manifold, but so are the challenges, for example with regard to data protection, the required database or integration into existing processes.

Two success factors are decisive: firstly, a clear definition of the problem to be solved, secondly, a step-by-step approach. The focus should be on relevant use cases rather than broad diversification.

‍

Comply with legal requirements: The EU AI Act

The EU AI Act has been in force since August 2024 - the world's first comprehensive law to regulate AI. Implementation is taking place on a phased basis, and there are already important obligations:

Banned since February 2025: AI systems with unacceptable risks (manipulative systems, social scoring, real-time biometric monitoring in public spaces).

Mandatory since February 2025: Companies must ensure that their employees have sufficient AI skills (Article 4 AI Act).

From August 2026: Comprehensive requirements for high-risk AI systems (for example in human resources, education, product safety).

Check at an early stage which AI systems you want to use and classify them according to risk categories. Make sure your teams are trained and document your processes.

The good news: The AI Act creates clear rules and can be seen as an opportunity to use AI responsibly and safely. Companies that act early on minimize risks and can take advantage of strategic advantages.

‍

Conclusion: AI strategy for SMEs — a real opportunity

For many companies, the focus is shifting. From the question of whether AI should be used to the question of where and how to get started.

The majority of companies already have an AI strategy and are planning increasing investments. At the same time, they barely exploit the existing potential. This is an opportunity for SMEs to achieve short-term success through process automation, more efficient use of data and more well-founded decision-making bases.

A sustainable AI strategy starts with a precise problem definition: Which specific challenge should be solved? Focused pilot projects, systematic learning and gradual competency development is an approach that has proven effective in practice.

You can find out how other SMEs have actually taken this path on the data:unplugged festival 2026 on March 26 & 27 in Münster. Here, companies from the e-commerce, industrial, retail, production and logistics sectors share their self-implemented use cases on AI strategies, process automation and AI integration: from finance to marketing, from IT to legal. At Mittelstandstage and a further four stages, we are creating space for exchange on well-founded practical examples of data and AI in order not only to understand the benefits of AI technology, but to actively shape them.

A successful AI strategy affects all areas of the company: For effective implementation, it is crucial to involve key people in your company, train them and positively prepare them for use. data:unplugged stands for a broad and well-founded transfer of knowledge — from which the entire business team benefits. Get a ticket for yourself and your core team now!