Navigating uncertainty in the presence of negative word of mouth: how different consumer modes impact innovation diffusion

Negative word of mouth (NWoM) can be a major factor in new product diffusion, as indicated by Goldenberg et al. (2007). We extend their work by accounting for the role of consumer uncertainty for NWoM’s impact on adoption dynamics. Using an agent-based simulation framework, we compare three different models for consumers’ possible navigation of uncertainty about the product quality—namely, the Goldenberg approach of ignoring uncertainty, Bayesian updating, and a novel approach (“Contradicting Signal Treatment Approach”, CoSiTA), in which consumers remain uncertain when facing contradictory information and delay their decision until the uncertainty is largely resolved. An empirical survey among German consumers suggests that all three types of behavior can be found in the population. Extensive computational experiments indicate that the Goldenberg model overestimates the detrimental effect of NWoM on product diffusion compared to the other two approaches. Furthermore, the diffusion under CoSiTA is considerably slower compared to Bayesian agents and generates a saddle in the adoption curve for certain social network structures. Simulations of a mixed market with all three types of consumers show that the interaction of groups who navigate uncertainty differently generates effects that qualitatively differ from a linear aggregation of the effects within each group.