Computational Consumer Behavior Modeling

Psychologists and researchers have studied how consumers respond to various presentations of products, information, and services. They are often able to describe how consumers operate in these situations, in simple rule-like descriptions. However, these consumer behavior descriptions are at the level of the individual, and they do not tell us what will happen if many consumers operate simultaneously on the same rule, or set of rules. By embedding these rules in computational agents, and then allowing the agents to buy and sell products in a virtual market, we can understand the collective result of individual consumer behavior models.

For instance, building on work by Arthur, Rand and Stonedahl have recently shown that as we increase the amount of time that individuals are allowed to optimize a decision the less efficient the result is for the group. In the future, agent-based modeling could be combined with participatory simulation to provide real data about how humans behave when presented with complex decision making tasks. By building these kinds of models, the project will facilitate a rich understanding of group consumer behavior.