A multiple ownerships diffusion model of cellular service: A study of the Swedish market

Other conference contribution, 2011

Cellular service diffusion has been widely explained using classic diffusion modeling which uses the accumulated number of subscriptions or possession of active SIM cards -often known as the cellular phone penetration rate- as the proxy. Such approach tends to overlook the presence of cellular service multiple-subscriptions as well as the problematic inaccuracies of cellular service penetration rate as the proxy. As an alternative, the present study explains the diffusion of cellular service using multiple ownerships diffusion models with longitudinal cross-section data as the proxy. Based on curve-fit estimation of the model using data obtained from nation-wide survey during 2002-2010, this study found that diffusion of cellular service with single subscriptions have higher innovation coefficient but lower imitation coefficient than one with multiple subscriptions. This indicates that the nature of demand for additional subscriptions may be different than that for first subscriptions and therefore the trend of growth should be estimated separately. Further, the multiple-unit ownership diffusion modeling offers more advantages than classic diffusion modeling as it can indicate the percentage of cellular service non-adopters in a population which can be estimated by examining the trend of cellular service adopters with single subscriptions to the population trend. The model also has a more intuitive forecasting power as it also distinguishes the growth trend projection for each type of adopters.