Linking Consumer Opinions with Reservation Prices in an Agent-Based Model of Innovation Diffusion

Abstract

We extend a recently developed agent-based model of innovation diffusion by linking the opinions of potential consumers with their market behavior via the concept of reservation prices. Through a dynamic mechanism that takes into account social influence, the agents in our model can both increase or decrease their product appraisal. Considering complete graph network structures and using mean-field treatment we find that the model can exhibit a plethora of scenarios, observed empirically but not attainable within the classical Bass model. We also show the existence of a critical market price above which the innovation cannot diffuse.

Impact of Social Interactions on Demand Curves for Innovative Products

“…Namely, one MCS consists of N elementary time steps ∆t in which one spinson is randomly selected and undergoes the procedure visualized in Fig. 1 in [5]. To consider only the stationary states of the system we collect the data after 10 3 MCS.…”

“…On the contrary, approaches utilizing opinion dynamics are able to reproduce the empirical data and are suitable for innovation diffusion modeling. In this study we take the latter findings as a staring point and use an opinion formation model proposed by Kowalska-Pyzalska et al [5] to investigate how the micro-level social interactions influence the shape of the demand curve for innovative products and producer decisions. By doing so, we fill the existing gap and contribute to the literature on innovation diffusion.…”

“…We utilize a relatively simple, yet sufficiently rich agent-based model developed recently by Kowalska-Pyzalska et al [5], which links the opinions of potential consumers † with their market behavior via the concept of the reservation price -P i , i.e. the highest price that agent i will accept and still purchase the good [7].…”

“…In order to comply with the commonly accepted standard in the economic literature, we present the results in terms of the demand curve, which considers the market price of the product as a function of the quantity sold, P (Q). However, we derive the demand curve from the penetration rate or level [2,3,5], Q(·), i.e. the percentage of adopted consumers as a function of model parameters (in particular, of market price P ).…”

“…In the adopted here so-called situation approach [16], each spinson behaves independently with probability p and with probability (1 − p) is exposed to WOM from a group of q neighboring (i.e. directly connected; in this study we use q = 4, for a discussion see [5]) spinsons or advertising. When a spinson is confronted by a unanimous group of q neighboring spinsons sharing the opposite opinion, it will conform to peer pressure [10] and choose a new reservation price (resulting in opinion change).…”