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 593

or insignifi cant (e.g., Currie and Park 2002 ; Leefl ang and Wieringa 2010 ;
Montgomery and Silk 1972 ; Parsons and Vanden Abeele 1981 ; Rosenthal et al. 2003 ).
Vakratsas and Kolsarici ( 2008 ) argue that the low level of marketing effectiveness
may be due to saturation effects.
A second important observation from Table 20.1 is that the effectiveness of
promotional expenditures exhibits a large degree of heterogeneity (Leefl ang and
Wieringa 2010 ; Manchanda et al. 2005 ; Narayanan et al. 2004 ). The larger number
of studies at the brand level also allows the discussion of possible moderators in
greater detail. We distinguish four different categories of moderators: (1) the type of
promotional instrument, (2) product category characteristics, (3) market character-
istics, and (4) model/data characteristics (Kremer et al. 2008 ).
Ad ( 1 ) Moderating effects of type of promotional instrument. Montgomery and Silk
( 1972 ) fi nd that brand level sales effectiveness differs by instrument and they rec-
ommend the inclusion of disaggregated promotional tools. Almost all later studies
that accommodate instrument-specifi c differences confi rm this heterogeneity in
effectiveness. The general fi nding is that detailing is the most effective promotional
tool, whereas DTCA is least effective at the brand level (Fischer and Albers 2010 ).
The effectiveness of other instruments such as journal advertising, meeting expen-
ditures, and direct mail is generally found to be lower than that of detailing, but
higher than DTCA. More generally, the group of direct to physician (DTP) instru-
ments has a stronger infl uence on demand than DTCA (Narayanan et al. 2004 ).
These outcomes correspond with Kremer et al. ( 2008 ) who estimate the following
instrument-specifi c elasticities: 0.32 for detailing, 0.123 journal advertising, and
0.073 for DTCA. The detailing elasticity estimate is remarkably close to that mea-
sured in a meta-analysis of personal selling by Albers et al. ( 2010 ) who report an
average detailing elasticity of 0.31.
There are several authors who fi nd that effect duration differs by promotional
instrument (e.g., Dekimpe and Hanssens 1999 ; Leefl ang et al. 1992 ; Montgomery
and Silk 1972 ). Consequently, it is important not only to accommodate lagged
effects for the different promotional instruments but also to allow for heterogeneity
in their dynamic response.
Some authors have investigated interactions between pharmaceutical promotion
instruments. Parsons and Vanden Abeele ( 1981 ) fi nd a negative interaction between
samples and printed information that sales representatives leave at a physician's
offi ce after a sales call. Such a negative interaction effect is sometimes referred
to as “jamming” (Azoulay 2002 ). Narayanan et al. ( 2004 ) fi nd signifi cant positive
interactions between detailing and DTCA, where, interestingly, DTCA effectiveness
benefi ts more from detailing efforts than vice versa. They fi nd that other marketing
activities (journal advertising, meetings, and events) impact negatively on DTCA
and detailing effectiveness.
Ad ( 2 ) Moderating effects of product characteristics. The studies in Table 20.1
suggest that product characteristics might infl uence promotional effectiveness.
Shankar et al. ( 1999 ) fi nd signifi cant order of entry effects: marketing effectiveness
reduces when a brand enters the market later. Rao and Yamada ( 1988 ) fi nd that