Large-Scale Agent-Based Models in Marketing Research: The Quest for the Mythical Free Lunch
International Journal of Business and Economics Research
Positioned as an alternative to equation-based methods, agent-based modelling (ABM) has shown notable promise in dealing with cases where the latter has proven inadequate. One of the areas where the limitations of traditional approaches are most pronounced is that of consumer behaviour research. A primary trait encountered within this scope is that of adaptability, and agent-based methods appear to be ideally suited to the task of capturing this dimension. It therefore follows that marketing
... s that marketing researchers are likely to gain novel and extensive insight by way of constructing large-scale, complex ABMs. However, the computational cost of complex simulations can be prohibitive. Furthermore, the literature makes little effort to elucidate means of making such ABMs feasible, beyond relying on natural hardware evolution. Unfortunately, the latter source of growth has grown stagnant, and the only avenue for the continued expansion of performance appears to be the move to parallel platforms and programming. The present research presents a cross-section of the current state-of-the-art in high-performance ABM frameworks, and proposes a novel approach to levering the as of yet untapped potential of cheap, ubiquitous Graphics Processing Units (GPUs). This insight is mapped into the space of consumer behaviour research, and a consistent argument is made in favour of larger, more detailed, ABMs, both as alternatives to current approaches as well as a development of prior forays into this area. In conclusion, a call to action is formulated, both to marketing researchers as well as computational economists, emphasizing the interdisciplinary requirements of ABM usage in the field of marketing.