The Bass Model In Action

The United States Department of Energy in 1980 used the Bass model to forecast the adoption of solar batteries. The DOE used a survey of home builders to aid in their initial choices for p and q values. Using these empirically suggested values, the DOE concluded that solar battery technology was not sufficiently robust to encourage word-of-mouth propagation. Because of their finding they postponed their proposed wide scale introduction of the technology until solar battery technology had improved to the point that new users would be satisfied with the technology and thus the higher q value would predict faster sales growth.

A decade later in the 1990s DirecTV had planned a launch of its subscription satellite television delivery service. Prudently, it attempted to obtain a prelaunch forecast for five years into the future. DirecTV’s forecast was again based on the Bass model and the p and q values were also obtained from a survey of prospective users; this information was combined with histories of similar services. The forecasts produced in 1992 were quite good from the company’s point of view and, after the fact, the estimates compared favorably with the actual 1994 to 1999 experience.

Numerous other firms have reported productive results using the Bass model. RCA in the mid-1980s used a modified Bass model to forecast the sales of music CDs as a function of CD player sales. The model proved quite accurate. The Bass model is also used routinely to predict box office revenues for movies and to make decisions on how many screens to use for a particular movie.

Source: Van den Bulte, Christophe (2002), “Want To Know How Diffusion Speed Varies Across Countries and Products? Try Using a Bass Model.” PDMA Visions, 26 (4), 12-15.