Hybrid Temporal-Graph Models for Predicting and Preventing Productivity Paradoxes in Service Operations

Olekanma et al. (2024) identify a paradox: productivity KPIs in financial institutions often drive unintended, paradoxical outcomes due to contextual differences in HR practices. Meanwhile, Sawant (2025) demonstrates the power of temporal-graph neural networks to model operational interdependencies in oil and gas. By uniting these threads, this research would construct a hybrid temporal-graph model that ingests HR, process, and outcome data in service industries (like banking or logistics) to forecast when productivity-driven interventions may backfire. The model would highlight specific organizational contexts where managerial actions (training, incentives, new KPIs) are likely to yield positive or negative effects, supporting more nuanced, data-driven management decisions.

References:

1. The dilemma of employee productivity measures and managerialism practices: an empirical exploration in financial institutions. Obafemi Olekanma, Christian Harrison, Adebukola E. Oyewunmi, Oluwatomi Adedeji (2024). International Journal of Productivity and Performance Management.
2. Compqual-Tgnet: A Novel Hybrid Temporal-Graph Neural Architecture for Analyzing Competency and Quality Metrics in Oil and Gas Operations. Shashank Sawant (2025). International Journal For Multidisciplinary Research.