Copyright
Glen L Urban.

Pre-test-market models : validation and managerial implications online

. (page 2 of 2)
Online LibraryGlen L UrbanPre-test-market models : validation and managerial implications → online text (page 2 of 2)
Font size
QR-code for this ebook


Decision makers must be careful in managing the enthusiasm for particular

new products. Brand managers personally may have much to gain and little to

lose by pushing their product even in the face of results that are below the

cutoff values. The usual persuasion methods after rely on proposing changes

that will be made to improve the product, identifying unrealistic conditions in

the test market (e.g. competitor's high sampling and promotion), or arguing

against the validity of the forecasting models. V/hile these sometimes may be

legitimate responses, care must be exercised not to drop the screening criteria

- 20 -



It has been assumed in the above analyses that fie reward function is not
changed by the test market. However, if the firm cannot produce the product in
national volumes at its specified level of quality, the estimated reward
obviously will not occur. Before considering elimination of test market, the
firm must be sure it can produce the product within the design tolerances. If
they are not sure, a test market may be required in any case. Another manner in
which the test market can affect the reward function is by improvement of the
product based on in-market experimentation and model analysis (see Urban and
Hauser). In this case, test marketing may be advisable because it increases the
expected reward. An example might be found in case 2 of Table 9, where if the
test market improved the brand's profit by more than $320,000, a test market
should be included in the sequential testing system. The advent of pre-test
market models implies that test market designs should allocate increasing
emphasis to finding the best marketing mix for a product rather than on sales
forecasting alone.



- 22 -



AO\NOWLEDQ'1ENTS
Special thanks to Steve Shugan who helped formulate the sequential
decision theory model and John Hauser and Leigh McAlister who made valuable
criticisms and suggestions. Thanks to Management Decision Systems, Inc. for
supplying the data for this study, to Barbara Petroulis, Mary Lyons, and
Robert Wallace who made major contributions to its collection, and the
cooperating companies who responded to the questionnaire.



- 24 -



5~ 'r- °)d



Date Due






Lib-26-67



MIT LIBRARIES



3 ^DflO 0DSfl30M3 2





2

Online LibraryGlen L UrbanPre-test-market models : validation and managerial implications → online text (page 2 of 2)