The dynamic market potential is nearly always a latent factor that drives the commercial performance of a product or service, namely, its diffusion. The modulation over time of a market potential may be grounded on the evolution of awareness of the product properties through a latent network among agents. This network may have different levels of connectivity, seeding effects and some type of heterogeneity of agents that affects their relationships. The GGM (Guseo and Guidolin (2009) [9]) introduces a specific evolutionary distribution for the market potential. This is based on the strong assumption of complete connectivity of the hidden network, supporting the growth of awareness focussed on specific pharmaceutical drugs or products with wide communication investments. Conversely, the basic idea proposed here is grounded on a convex combination of the Fibich–Gibori distribution, obtained for a minimally connected one-dimensional (1D) network topology, with the Bemmaor–Lee distribution, which takes into account unobserved heterogeneity aspects of agents in a market under a complete connectivity. Based on a continuum between opposite extremes, the extended final model, the Network Automata GGM (NA-GGM), which includes the GGM as a special case, allows the modulation of the involved latent network by exploiting the observed time series of sales and avoiding rigid assumptions on the latent network topology. A specific application of the new model is discussed in detail in term of the weekly diffusion of a statin, Rextat, in the central part of Italy. The proposed extension, NA-GGM, is statistically significant concerning the GGM and it is more coherent in commercial behaviour forecasting.
Modelling dynamic market potential: Identifying hidden automata networks in the diffusion of pharmaceutical drugs
Guseo R.
Formal Analysis
;
2021
Abstract
The dynamic market potential is nearly always a latent factor that drives the commercial performance of a product or service, namely, its diffusion. The modulation over time of a market potential may be grounded on the evolution of awareness of the product properties through a latent network among agents. This network may have different levels of connectivity, seeding effects and some type of heterogeneity of agents that affects their relationships. The GGM (Guseo and Guidolin (2009) [9]) introduces a specific evolutionary distribution for the market potential. This is based on the strong assumption of complete connectivity of the hidden network, supporting the growth of awareness focussed on specific pharmaceutical drugs or products with wide communication investments. Conversely, the basic idea proposed here is grounded on a convex combination of the Fibich–Gibori distribution, obtained for a minimally connected one-dimensional (1D) network topology, with the Bemmaor–Lee distribution, which takes into account unobserved heterogeneity aspects of agents in a market under a complete connectivity. Based on a continuum between opposite extremes, the extended final model, the Network Automata GGM (NA-GGM), which includes the GGM as a special case, allows the modulation of the involved latent network by exploiting the observed time series of sales and avoiding rigid assumptions on the latent network topology. A specific application of the new model is discussed in detail in term of the weekly diffusion of a statin, Rextat, in the central part of Italy. The proposed extension, NA-GGM, is statistically significant concerning the GGM and it is more coherent in commercial behaviour forecasting.
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
