The Allure of the Grand Scale

Many SMEs look to examples from large corporations with large language models, company-wide chatbots, and AI strategies with their own roadmaps. That sounds impressive, but it rarely fits the reality of a company with 80 or 300 employees. The resources for complex transformation projects are missing or not available, and often neither is the data set on which large models truly shine.

Where the Real Benefits Lie

The actual economic leverage for small and medium-sized businesses almost always lies in the small details: in recurring, clearly definable inefficient tasks that are time-consuming and involve little strategic complexity. These include, for example:

• Document processing: Automatically capturing and verifying invoices, delivery notes, or contracts
• Customer communication: pre-sorting standard inquiries, drafting responses, and expediting translations
• Knowledge management: Making internal know-how searchable to ensure that information is readily and fully available and that experience-based knowledge is not tied to individual employees
• Quotation and pricing processes: Intelligently compiling templates and text modules

These use cases are unspectacular, but that is precisely what makes them economically attractive. They can be implemented individually with reasonable technical and time investments, their benefits are measurable, and they do not tie up critical resources for months on end.

The Difference Between Efficiency and Transformation

A second point to consider: AI in small and medium-sized businesses is often positioned as a transformation topic, but it usually first demonstrates its value as an efficiency tool. Not every company needs to reinvent its business model. Far more often, the goal is to make existing processes significantly faster, less error-prone, or more scalable. This modesty in goal setting is not a weakness, but often the reason why projects are actually completed rather than fizzling out in the pilot phase.

A Top Management Priority, Not Just an IT Project

A common structural flaw intensifies the problem: AI implementation is delegated to the IT department under the assumption that “This is technology, let the experts handle it.” This is precisely where the misunderstanding lies. IT can integrate systems, ensure data protection, and build interfaces. But it cannot decide which processes are strategically important, where time losses are most painful, or which customer relationships are particularly sensitive to automation. These are business questions, not technical ones.

When management treats AI as an IT issue, one of two things often happens: Either technically sound but isolated solutions emerge without any real economic impact, because no one with a holistic view has set the priorities, or the initiative fizzles out because it lacks the backing needed to overcome departmental boundaries, since AI projects almost always affect multiple areas simultaneously: sales, customer service, production, accounting, …

Successful medium-sized companies therefore treat AI implementation as a leadership task.

Top Management defines first which business problem needs to be solved, prioritizes competing ideas from different departments, and bears responsibility for financial success. IT remains a key implementation partner—but just that: a partner, not a client.

The Real Prerequisite: Clarity Before Technology

The biggest stumbling block is rarely the technology itself. It is the lack of clarity about which specific problem needs to be solved and how success will be measured. Companies that first precisely describe a task—Who is doing this today? How long does it take? What does a mistake cost?—before discussing tools have significantly higher success rates than those that start with the question, “What kind of AI do we need?”

This goes hand in hand with the need to adapt the organization, engage and involve teams and employees at all levels, alleviate fears and reservations, and foster a positive outlook toward the future.

Initiation and assurance of the consistent implementation of these efforts is a crucial responsibility of senior management, not for IT.

Conclusion

In small and medium-size enterprises, economic benefits from AI don’t come from grand visions, but from consistently solving small, clearly defined problems and from leadership that actively shapes this process rather than delegating it. Those who follow this path not only achieve quick successes but also build the organizational competence needed to take bigger steps later on, with realistic expectations and a keen eye for where technology truly makes a profitable difference.