A simple algebraic estimation procedure for innovation diffusion models of new product acceptance

Abstract

A simple algebraic estimation procedure is developed to estimate the parameters of diffusion models of new product acceptance. The procedure required knowledge of the occurrence of the point of inflection (based on actual data, analogous products, or management judgments). It is conceptually easy to use and can be implemented by using a hand calculator. Since the procedure does not employ period-by-period time-series diffusion data, it is not expected to provide the best fit to the data as compared to the maximum likelihood and nonlinear least squares estimation procedures. However, the procedure can provide very reasonable estimates about the relative magnitudes of the parameter estimates. In that respect, the procedure can be used to generate good starting values for the maximum likelihood and nonlinear least squares estimation procedures. In the absence of data, using management judgments about the point of inflection, the procedure can be implemented in a decision-support system to develop conditional diffusion curves for a new product. Data from four diverse innovations are used to illustrate the procedure.