Soutenance de thèse Negin Safaei

• Prof. Kazem HAKI, co-directeur de thèse, Haute école de gestion de Genève 
• Prof. Jean-Henry MORIN, co-directeur de thèse, Université de Genève 
• Prof. Giovanna DI MARZO SERUGENDO, président du jury, Université de Genève 
• Prof. Stephan AIER, Université de Saint-Gall 
• Prof. Katarzyna WAC, Université de Genève

This dissertation examines how service innovation is coordinated in B2B platform ecosystems, focusing on two interrelated developments: the integration of Generative Artificial Intelligence (GenAI) into enterprise platforms and the role of incumbent-led B2B innovation platforms in shaping partner innovation behavior and outcomes. It is structured into two complementary research streams that address these developments from ecosystem-level and platform-level perspectives.

The first research stream investigates GenAI integration into enterprise platforms using single case study. It develops a process theory explaining how enterprise platforms adapt through architecture-governance reconfigurations. The findings show that GenAI introduces a logic of probabilistic inference, which challenges the existing logic of determinacy embedded in enterprise systems. This destabilization affects both architectural mechanisms (e.g., stable interfaces, extension containment) and governance mechanisms (e.g., certification, monitoring). Platform owners respond through coordinated adjustments that enable the coexistence of deterministic and probabilistic components. This results in a hybrid platform configuration, in which determinacy remains the primary organizing logic while probabilistic capabilities are integrated through new architectural and governance mechanisms. These reconfigurations give rise to emergent platform properties, including increased generativity, boundary fluidity, and trust mediation.

The second research stream examines incumbent-led B2B innovation platforms using a simulation-based theory development approach. It explains how service innovation outcomes emerge from the interaction between the complexity of market opportunities and platform mechanisms, specifically transaction mechanisms and production mechanisms. The results show that partners’ decisions to innovate on-platform or off-platform depend on whether platform mechanisms provide sufficient advantages in coordinating multi-partner collaboration. The findings further demonstrate that generativity and profitability are distinct and may diverge, particularly under conditions of high complexity and intensified competition. Moreover, the outcomes differ between platform owners and partners, highlighting asymmetries in value creation and value capture within B2B ecosystems.

By combining these two streams, the dissertation shows that service innovation in B2B platform ecosystems is shaped by the joint effects of ecosystem complexity and technological complexity. It provides a theoretically grounded explanation of how platform architecture, governance mechanisms, and partner behavior interact.