Abstract
As microservice systems evolve, their architectures often reflect team collaboration patterns. Cross-service developer behaviors introduce Organizational Coupling (OC), which may lead to architectural degradation. The link between such behaviors and degradation remains underexplored. This paper investigates whether OC predicts architectural degradation in microservice systems. A pilot study analyzes architectural evolution and organizational behavior data from Spinnaker. Granger-causality tests with Benjamini-Hochberg correction are applied to architecture health metrics, including volatility in service cohesion and spikes in service coupling, with validation through a reverse-direction placebo test. Results indicate that spikes in OC consistently precede cohesion volatility and coupling hotspots, establishing a directional link between suboptimal organizational behaviors and architectural degradation. Monitoring OC, therefore, provides a simple, data-driven signal to anticipate architectural risk and guide preventive actions in microservice systems. All materials are released in a replication package.