Estimating Contagion on the Internet: Evidence from the Diffusion of Digital/Information Products

Even as the Internet continues to grow as a global platform for communication and commerce, the success of new value offerings on the Internet hinges on acquisition of new customers and retention of existing customers. Central to the flow of customers in and out of trial and repeat behavior in this burgeoning and dynamic environment, characterized by diversity among both producers and consumers of value offerings, is the process of social contagion—active word of mouth that flows among customers or passive observation of others. To estimate contagion on the Internet, the authors develop a trial-repeat purchase diffusion model for successive innovations in value offerings on the Internet. The model extends the state-of-the-art diffusion modeling by incorporating (i) dynamic market potential, (ii) heterogeneity among first-time triers, (iii) heterogeneity in word of mouth due to repeat buyers and non-repeaters (i.e., positive and negative word of mouth), and (iv) dynamic repeat purchase rate. The model also incorporates the influence of product characteristics, specifically source of innovation (i.e., whether the innovation is driven by environmental needs or competitive pressures) and product bundling, and competition. The authors test the model with weekly adoption data for 11 computer software products available on the shareware system, involving over 100 new versions in the period 1991–1994, and in a market whose size grows by a factor of fifty from early 1991 to late 1994. The findings clarify the role of word of mouth effects, competition, and product characteristics in fostering the diffusion process for digital information goods.