AI-Driven Analysis to Extend and Augment Qualitative Research on Non-Linear Firm Growth

Abstract

Firm growth has been a central topic in the entrepreneurship literature for decades (Penrose, 1959), yet dominant stage models have faced sustained criticism for their empirical limitations (Phelps et al., 2007; Brown et al., 2017). Building on a conceptualisation of firm growth as a non-linear, multidimensional phenomenon, this paper revisits an established qualitative research project on "entrepreneurial leaps" as transitional growth phases between dynamic organisational states (Sternad & Mödritscher, 2022; Levie & Lichtenstein, 2010). The original study employed a multiple-case design with in-depth interviews conducted with leaders from 24 companies across three industries in Austria, Germany, and Italy.The present contribution applies three state-of-the-art large language models, namely ChatGPT, Google Gemini, and Claude, to re-analyse the existing and extensive dataset. Such a comparative reanalysis pursues three interrelated objectives: (1) examining whether established patterns are corroborated or nuanced through machine-driven interpretation, (2) surfacing emergent patterns not previously foregrounded in the human-led analysis, and (3) investigating potential differences in firm growth dynamics across the three industries under study.Rather than positioning AI as a substitute for human interpretive expertise in science, this experimental study treats it as a methodological augmentation that might extend the analytical reach of the original project. The paper concludes by deriving a comprehensive research agenda, grounded in human scholarly judgment. Therefore, it contributes to the emerging discourse on hybrid methodologies in qualitative research and invites reflection on the role of AI in qualitative inquiry.