Journalpublished

Business Dynamics: Systems Thinking and Modeling for a Complex World

John D. Sterman (MIT Sloan School of Management)
T
Curated by Ted Lango
Published May 9, 2026Updated May 10, 2026
Read Original ↗

Abstract

This comprehensive textbook provides the foundations of system dynamics for modeling complex business and social systems. It covers causal loop diagrams, stock-flow models, feedback structures, and simulation, with extensive applications to operations, strategy, and public policy. Sterman identifies ten properties that characterize systems exhibiting counterintuitive behavior.

Curator Summary

Sterman's ten properties of complex systems — constantly changing, tightly coupled, governed by feedback, nonlinear, history-dependent, self-organizing, adaptive, characterized by trade-offs, counterintuitive, and policy resistant — describe contact center automation perfectly. This book provides the formal methodology for modeling the reinforcing feedback loops that create second-order effects: the latent demand loop, the complexity concentration loop, and the emergent inquiry loop. If you're trying to explain to leadership why 'obvious' automation investments produce counterintuitive results, Sterman gives you the framework.

Why It Matters

System dynamics explains why linear thinking about automation fails. The three reinforcing loops in contact center AI — demand rebound, complexity concentration, emergent inquiries — are textbook system dynamics patterns. Sterman's stock-flow formulation provides the mathematical structure for modeling these interactions: stocks like Active Customer Base, Contact Propensity, and Complexity Mix; flows like Demand Generation Rate and Savings Erosion Rate. This is how you build a simulation model that captures reality.

Caveats

This is a general systems dynamics text, not specific to contact centers or WFM. The application to service operations requires domain expertise to construct appropriate models. System dynamics models can become arbitrarily complex — the challenge is finding the right level of abstraction. Simulation results are only as good as the parameter estimates fed into them.

Discussion

No comments yet. Start the discussion.

Loading...