Exploring Productivity Gains from GenAI-Enabled Virtual Assistants: A Configurational Study of a Multifaceted Phenomenon

Enterprises increasingly embed generative AI–enabled virtual assistants (VA) into enterprise productivity software to improve digital work and performance. Evidence indicates that productivity gains (PG) are a multifaceted phenomenon of organized complexity shaped by interdependent organizational, technological, and human factors. Drawing on established IS lenses, this paper develops a theory-driven model of when and how VA-enabled enterprise productivity software yields PG. This paper deductively derives a research model specifying four IS theory-based conjunctural factors, and articulates equifinal pathways that enable or inhibit PG. To evaluate necessity and sufficiency, the paper outlines an expert interview-based fuzzy-set Qualitative Comparative Analysis research design. The contributions are threefold: (i) defining VA-enabled PG within enterprise productivity software as a multifaceted phenomenon; (ii) introducing a configurational research model; and (iii) extending IS business-value theorizing to AI-augmented digital work through set-theoretic reasoning and theory-driven propositions for empirical validation. The model clarifies when VA enhance enterprise performance and offers a testable roadmap for research on human–AI collaboration.