Service Diffusion Models: A case study on an Italian Travel Start-up

The Bass model of diffusion has been widely applied to forecast the diffusion of products prior to launch; instead, studies on services diffusion are still at the embryonic stage. This work presents an empirical analysis of the diffusion of an Italian travel start-up. Firstly, analogy was used to find similar services to the analyzed one in order to estimate the Bass parameters by simulating a case in absence of data. These parameters were then compared to the ones derived by NLR (Non-Linear Regression) techniques. Finally, being the sales history of travel agencies strongly characterized by seasonality, the model developed by Guidolin and Guseo (2012) was applied. Results show that the analysis by analogy might be biased, because of the limitations in data availability and the many assumptions required. The Bass model with seasonality instead, outperforms the Bass model in providing forecasts.