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 599

where only the adoption process is of interest. Later extensions of the Bass model
also accommodate repeat purchases; these models are known as trial-repeat diffu-
sion models.

20.3.2 Bass Framework Extensions: Pharmaceutical

Marketing Variables

Diffusion models are commonly criticized for their implicit consideration of the
effects of marketing variables (such as price and advertising) on the diffusion
process. Explicit inclusion of marketing variables in diffusion models not only
provides a better description of reality but also provides suggestions for altering the
diffusion process by manipulating the marketing variables. The importance of
including marketing mix variables in diffusion models has been highlighted by
several researchers (see, e.g., Bass et al. 2000 ; Bass et al. 1994 ; Mahajan and Muller
1979 ; Mahajan and Wind 1986 ). In this subsection we review several empirical
applications of macro-level diffusion models in the pharmaceutical industry and
examine which marketing variables they include and in which part of the diffusion
model the variables are included.
We distinguish several approaches to the inclusion of marketing variables in dif-
fusion models. Firstly one can incorporate the effects of marketing variables into the
size of the potential market (see, e.g., Berndt et al. 2003b ) which is comparable to a
market expansion effect (see Sect. 20.2.1 ). Another approach accounts for the effect
of marketing variables on the probability of adoption . The literature points to three
possible ways for marketing variables to affect the adoption rate:

(a) As an external infl uence (see, e.g., Hahn et al. 1994 ; Lilien et al. 1981 ; Rao and
Yamada 1988 ; Ruiz-Conde et al. 2011 ) where marketing variables affect the
adoption decision of a potential adopter without being infl uenced by informa-
tion from an early adopter.
(b) As an internal infl uence (see, e.g., Ruiz-Conde et al. 2011 ) where marketing
variables stimulate interpersonal communication. For example, the internal
infl uence of price should be perceived as a process where the decision of non-
adopters is affected by social interaction with adopters concerning (fl uctuating)
prices of the new product.
(c) As a mixed infl uence in which case marketing variables affect the trial rate
jointly through external and internal infl uence (see, e.g., Ruiz-Conde et al.
2011 ; Shankar et al. 1998 ; Vakratsas and Kolsarici 2008 ).

A fi nal way of incorporating marketing variables is to account for their effect on
the repeat rate (see, e.g., Lilien et al. 1981 ; Rao and Yamada 1988 ).
Most of the published studies reveal the importance of explicitly considering
the parameters of external and internal influences on the rate of adoption of
new products (Chandrasekaran and Tellis 2007 ). Extensions of the mixed influ-
ence diffusion models are recommended but are also more complex to estimate