Marketing Spending Models - Innovation and Marketing in the Pharmaceutical Industry - page 568

separately. Moreover, in most cases, the total budget is set by top management and
provides a financial restriction to all subsequent budgeting tasks. On the positive
side, we know that profit gains from optimized allocation in the amount of 40-60 %
by far outreach the gains from an optimized total budget that total only 3-5 %
(Mantrala et al. 1992 ; Tull et al. 1986 ).
The literature on normative models of pharmaceutical marketing spending is
emerging. We may broadly differentiate between static and dynamic models. Static
models try to maximize short-term profits and obtain optimal spend levels for the
current period. Dynamic models try to optimize the budget allocation over time and
may consider maximizing discounted future profits. The suggested approaches are
based on marginal analysis or simulation of market response functions that are part
of the profit objective function. The main focus is on detailing as largest pharmaceu-
tical spend category.

19.4.1 Static Optimization

Optimizing detailing expenditures across physicians . Manchanda and Chintagunta
( 2004 ) attempt to optimize the number of times a physician is detailed in a particu-
lar quarter. They model the number of prescriptions, y it , written by physician i in
quarter t to follow a Poisson distribution with a physician-specific mean λ it :

y

l exp( − l )

it

( ) =

it it

(19.8)

Pr yy | l

it = it it

y !

it

where

( )

2

l = exp b + b NDET + b NDET

it 0 i 1 i it 2 i it

NDET it denotes the number of times a physician is detailed in quarter t and ß is
a parameter vector to be estimated. Differentiating the mean prescription rate with
respect to NDET and setting zero gives the optimal number of calls per quarter a
physician should get. This point can be computed by − bb
1 i / (Manchanda and

Chintagunta 2004 ). The application of this model to a specific drug and a large
sample of physicians demonstrates that quite a large group of physicians is over-
detailed. By reallocating the wasted resources from over-detailed physicians to
under-detailed physicians in two different segments, prescriptions rise by 9 % and
11 %. Because the reallocated resources are evenly spread across the physicians, the
authors note that the revenue gain should be even higher for a more optimized rule.
Note that the total detailing budget remains constant. Thus, the increase in revenues
is fully profit-relevant. The improvement in profits should be even higher because
the reference base is lower than that for the revenue increase.

2

2