Theoretical Evolution of Dynamic Pricing in E-Commerce: from Cost-Plus to Data-Driven Paradigm Shift

Abstract

There is a fundamental theoretical evolution of dynamic pricing during the age of e-commerce, whereby traditional cost-plus methods are replaced by flexible, data-informed approaches that react to changes in market dynamics. This study will describe the history of dynamic pricing theory by highlighting the evolution of its theories through three main phases: those rooted in the cost-plus and marginalist theories in the early neoclassical era, followed by the behavioral and game-theory extensions in the late twentieth century and, finally, the machine learning revolution of our times. Using a synthesis approach of insights from various disciplines of economics, marketing science, computer science, and behavioral psychology, we develop a framework that reconciles algorithmic optimization with consumer welfare considerations. From this study, we find that while today’s paradigm of data-driven dynamic pricing leads to tremendous efficiency gains and maximized revenues, it also presents new dilemmas such as algorithmic collusion, price discrimination issues, and consumer perception of unfairness. We argue that future dynamic pricing theory should not only involve technical optimization but also consider other aspects like governance, regulation, and ethics. This study provides an integrated historical overview of the development of dynamic pricing ideas, the persisting trade-off between efficiency and equity, and an integrative research agenda for next generation of pricing theory.