Mapping Digital Transformation in Small and Medium-Sized Enterprises Through OpenAlex: A Bibliometric and Lexical Clustering Analysis

Digital transformation in small and medium-sized enterprises (SMEs) has expanded from a technology-adoption issue into a broader question of organizational capability, knowledge integration, market adaptation, and resilience.This study maps the development and conceptual structure of the OpenAlex-indexed literature on SME digital transformation from 2017 to 2025.The current cleaned corpus contains 755 articles and reviews.Descriptive bibliometrics were combined with keyword co-occurrence analysis, document-level term frequency-inverse document frequency and k-means lexical clustering, and co-authorship network analysis.The corpus shows a sharp increase in publication activity in 2024-2025, substantial dispersion across journals and authors, and a strong geographic concentration of Indonesian affiliations among records with usable country metadata.Keyword communities position knowledge management close to digital transformation and SME capability development, while five lexical research fronts emphasize micro, small, and medium enterprise financial literacy; digital adoption and knowledge management; performance, innovation, and sustainability; digital marketing and social media; and COVID-19 resilience.Co-authorship analysis further indicates a highly fragmented collaboration structure, with 2,260 authors distributed across 689 connected components and only 89 of 755 studies recording international co-authorship.The study contributes by distinguishing among concept-level keyword communities, document-level lexical research fronts, and collaboration structures within an openly reproducible OpenAlex corpus.The findings suggest that SME digital transformation is best interpreted as a knowledge-intensive capability reconfiguration process rather than as simple technology acquisition.Because OpenAlex coverage is broader and geographically different from that of curated databases, the results are database-sensitive and require crossdatabase and citation-based validation.