Pharmaceutical Product Market Share Estimation: Full-Factorial Attraction Model
The study objective is to test whether a newly proposed Full-Factorial Attraction Model allows for more precise estimation and predictability of market share when accounting for a new entrant. An Attraction Model is often used to predict market share, although its predictability power is limited to...
Gespeichert in:
Veröffentlicht in: | Marketing bulletin 2008-01, Vol.19, p.2 |
---|---|
Hauptverfasser: | , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | The study objective is to test whether a newly proposed Full-Factorial Attraction Model allows for more precise estimation and predictability of market share when accounting for a new entrant. An Attraction Model is often used to predict market share, although its predictability power is limited to a constant market structure (Cooper et al. 1996). Assuming a constant market structure across time does not solve the issue as the new entrant behavior has a direct impact on the incumbents (Fok 2003). Howie and Kleczyk (2007) proposed a joined pre- and post- new brand introduction model based on a re-conceptualization of any market share as a series of two-brand groups (Howie & Kleczyk 2007). In this investigation, both model types are evaluated on their goodness-of-fit and regression estimate stability for joined and separate pre- and post- new brand introduction models. OLS and GLS panel estimation is utilized in econometric modeling. The study results reveal that the standard way of estimating market share does not account for the changing market structure. On the other hand, the Full-Factorial Attraction Model does account for the changing market structure. The impact of the explanatory variable stays unchanged across all model specifications and no structural break is observed. The predictive power of the model is also the highest for the Full-Factorial Attraction Model. [PUBLICATION ABSTRACT] |
---|---|
ISSN: | 0113-6895 1176-645X |