Modeling the Effects of Promotional Efforts on Aggregate Pharmaceutical Demand: What We Know and Challenges for the Future - Innovation and Marketing in the Pharmaceutical Industry - page 603

between the two analyzed geographical areas, while the innovation adoption pattern
is essentially equal. They show that the higher value for the imitative infl uence
parameter in one area is the result of a more effective communication strategy.

20.3.3 Conclusions

Today the new interdependencies among consumers, such as word-of-mouth
communication, network externalities (especially, online social networks) and
social signals, new types of product category, and different kinds of consumer
behavior (such as multiple purchases of a new product by a single purchaser),
increase the complexity of the diffusion process of new products. This complexity
obliges researchers to develop and use new modeling approaches (e.g., Fok and
Franses 2007 ; Sood et al. 2009 ) and to continually adapt their methods of describing
and modeling these diffusion processes. This continuous development causes diffu-
sion modeling in marketing to remain an active research fi eld, as refl ected in the
more recent review papers on innovation diffusion (Frenzel and Grupp 2009 ; Meade
and Islam 2006 ; Peres et al. 2010 ). More research into, and reports of, the actual use
of diffusion models in marketing is an ongoing request in the literature. Although
the main application areas, in terms of practical impact, are consumer durables and
telecommunications (Meade and Islam 2006 ), the pharmaceutical sector is one of
the key industries of interest in today's new diffusion modeling efforts (Mahajan
et al. 1990 , 2000 ; Peres et al. 2010 ).
We now return to the initial questions posed: which are the pharmaceutical
marketing variables that are mainly investigated? And where and how should they
be included in the diffusion models? The most important and most researched
marketing variable is detailing and its effect on the diffusion process of prescription
drugs is always signifi cant and has the expected sign (i.e., positive for own detailing,
negative for competitive detailing), except in one study (where the authors present
several explanations to justify the lack of signifi cance). The infl uence of other
marketing variables such as medical journal advertising and physician meetings is
also investigated and their impact on the trial rate is revealed with the expected sign.
However, as pharmaceutical marketing activities are highly correlated (also reported
by Gatignon et al. 1990 ; Rizzo 1999 ) most studies combine them into an aggregated
variable. The effect of DTCA is also investigated in Ruiz-Conde et al. ( 2011 ) who
demonstrate its relevance on the diffusion process.
The studies reviewed here provide insights into where and how to include
pharmaceutical marketing variables into diffusion models. Only one study (Berndt
et al. 2003b ) considers the effect of marketing variables on the size of the potential
market. Three studies (Hahn et al. 1994 ; Lilien et al. 1981 ; Rao and Yamada 1988 )
assume that promotional efforts affect the diffusion process via external infl uence
and fi nd signifi cant corresponding coeffi cients. Two studies (Shankar et al. 1998 ;
Vakratsas and Kolsarici 2008 ) assume that promotional efforts affect the diffusion
process via both external and internal infl uences, and also fi nd signifi cant