Mantel, PeterSilva Barros , Diego OsvaldoBerlin International University of Applied Sciences2025-12-102025-12-102025https://repository.berlin-international.de/handle/123456789/1244The rapid growth of data in the digital economy has positioned machine learning (ML) as a transformative technology for improving decision-making, efficiency, and competitiveness. While large corporations have the resources to integrate ML through dedicated data teams and advanced infrastructures, e-commerce teams often face significant barriers. Despite generating valuable data through sales, customer interactions, and marketing, many teams struggle to adopt ML effectively due to limited resources, lack of expertise, and uncertainty about practical implementation. In 2024, only 11% of European SMEs reported using ML, highlighting a substantial adoption gap compared to larger firms. This thesis investigates how small e-commerce teams with no formal data science background perceive, explore, and begin to adopt ML tools. Drawing on Diffusion of Innovations theory and the Dynamic Capabilities framework, the study emphasizes both the human and organizational dimensions of adoption. A mixed-methods approach was employed, including structured interviews with team members at Berlin Brands Group, an educational workshop, and a live demonstration of ML applied to a practical dataset. This process allowed the researcher to capture changes in perceptions, readiness, and confidence before and after exposure to ML. The findings highlight three central themes: (1) adoption challenges at the technological, organizational, and personal levels; (2) enabling factors such as curiosity, ease of use, and peer examples; and (3) shifts in awareness and openness following experiential learning. The study contributes to both theory and practice by offering a grounded understanding of ML adoption in SMEs and by proposing practical recommendations to make ML more accessible, actionable, and democratic for small teams.Machine LearningE-commerceDiffusions of InnovationsDynamic CapabilitiesThesis