WEBU JournalSlovenčina

Research overview · 13.11.2026

How Small Businesses Can Adopt Generative AI Without Chaos

Team members discussing a project at a laptop
Illustrative photograph: Ivan S / Pexels · Pexels License

Author: Martina Mundová

Having an AI tool does not automatically create value. New research highlights the role of trust, managerial direction and the move from experimentation to repeatable processes.

How the research was conducted

Siljeur, Schulte and Kanbach carried out a qualitative study based on 31 semi-structured interviews with decision-makers involved in generative AI adoption by SMEs. Most participants were based in Germany and France, with further perspectives from South Africa and Vietnam.

The research did not calculate a universal financial saving. It explored how technological and organisational conditions shape adoption.

Three connected stages

The authors describe three phases: activating individual trust and engagement, legitimising adoption through leadership and direction, and finally embedding AI into routines to create sustained value.

Employee initiatives and managerial support need to work together. Isolated experiments may not become shared processes without clear organisational rules.

A practical WEBU approach

Start with one recurring task, such as drafting answers to common customer questions. Assign a person to review outputs, restrict personal data entry and specify quality criteria.

After a pilot, compare correction time, errors and employee feedback before scaling. These are WEBU suggestions, not interventions tested in the study.

Tools are not the whole system

AI tools enter existing relationships, working habits and decision structures. Training, sharing errors and concise documentation can therefore matter as much as choosing a model.

In website production, AI-generated text should remain separate from fact-checking, language editing and final publication approval.

Limits and conclusion

Qualitative interviews offer insight into participant experience but cannot guarantee a return on investment for every business or country.

The useful sequence is to build trust, clarify responsibility and then test whether AI delivers sustained value in real workflows.

Research source

Siljeur, J., Schulte, N., & Kanbach, D. (2026). Activate, legitimize, embed: enacting generative artificial intelligence adoption in small and medium-sized enterprises. EuroMed Journal of Business. https://doi.org/10.1108/EMJB-01-2026-0061

Author: Martina Mundová