An agent-based spatial model of consumer behavior
Unilever Corporate Research
Last modified: March 31, 2006
Understanding the drivers for consumer behavior and the underlying mechanisms of market dynamics is of key importance to marketing practitioners, market strategists and competition and regulatory bodies. Furthermore the cross-disciplinary nature of this area has attracted the interest of researchers in diverse fields such as economists, social scientists, psychologists and computational scientists. Marketers usually ask questions like “What are the chances for success when a new product is launched (e.g. brand extension)?” or “How does word-of-mouth affect sales and the long-term prospects of a brand?” To answer these questions, traditional marketing models, based on equilibrium statistics and macro variables such as market share and price elasticity, have usually been employed with limited success. As a powerful computational method, agent based modeling has the potential to address many of the shortcomings of traditional techniques. Specifically, ABM (i) allows the dynamic nature of markets to be modeled, (ii) treats consumers as individuals therefore retaining the richness of information at the micro level and (iii) allows consumer interactions and social networks to be explicitly modeled. Here we describe the a spatial agent based consumer behavior model drawing from the marketing and the behavioral sciences literature. We present some simulation results exploring and comparing different market diffusion models using real, individual-based market data.