Modelling Trade Promotion

  • View

  • Download

Embed Size (px)

Text of Modelling Trade Promotion

  • 8/6/2019 Modelling Trade Promotion


    Modelling the Effectiveness and Profitability of Trade Promotions

    Author(s): Robert C. Blattberg and Alan EvinSource: Marketing Science, Vol. 6, No. 2 (Spring, 1987), pp. 124-146Published by: INFORMSStable URL: .

    Accessed: 02/08/2011 17:48

    Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . JSTOR's Terms and Conditions of Use provides, in part, that unless

    you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

    may use content in the JSTOR archive only for your personal, non-commercial use.

    Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . .

    Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

    page of such transmission.

    JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

    content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

    of scholarship. For more information about JSTOR, please contact [email protected]

    INFORMS is collaborating with JSTOR to digitize, preserve and extend access toMarketing Science.
  • 8/6/2019 Modelling Trade Promotion


    MARKETING SCIENCEVol. 6, No. 2, Spring 1987Printed in U.S.A.


    University f ChicagoInteractiveData CorporationTrade promotions have become an increasingly important element of the marketing mix.Yet, there is very little researchdescribing how to measure the profitabilityand effectiveness oftradepromotions. This paperdescribes how retailers behave when tradepromotions are offered.

    Then, a model is developed to capture the two key components of the process: the consumerand the retailer. An example is given showing how to apply the model to actual manufacturerand retail sales data. Then estimates of the profitabilityfor different items in a product categoryare calculated.Many research questions are raised in this paper which can serve as future directions forresearch.Why are trade promotions generally unprofitable?How can scanner data improve theestimates given? How do different types of trade promotions affect the retailer and ultimatelythe consumer? Which brands and items should be trade promoted?

    (Modelling; Trade Promotion;Scanner Data)

    1. IntroductionBrand managers devote substantial dollar resources to trade promotions and, inmany companies, trade promotion budgets are much greaterthan advertising budgets.Yet, there have been very few models developed to measure the profitability and saleseffects of trade promotions.The purpose of this article is to present a model which assists management in under-standing and measuring the effects of trade promotions. The specific uses of the modelpresented are: (1) to evaluate individual promotions, (2) to identify the best tradepromotions for each size and in each geographical area, (3) to evaluate future promo-tional plans, and (4) to develop trade promotion tactics.The model developed is applied to a data set using Nielsen consumer sales data(bi-monthly) and company shipment data. The consumer sales data used have manyimitations which are discussed later. However, the general modelling approach can beapplied to scan data such as Nielsen's Scan Track. The model can actually be applied

    more easily to these data than the data described in the paper since more accurateconsumer promotions and weekly sales data are available from Scan Track.Before continuing, it is useful to define a trade promotion. Trade promotions arespecial incentive programs offered by the manufacturer to their distribution channelmembers. They take many forms, including direct price discounts and free case offers.124

    0732-2399/87/0602/0124$01.25Copyright ? 1987, The Institute of Management Sciences/Operations Research Society of America

  • 8/6/2019 Modelling Trade Promotion


    EFFECTIVENESS AND PROFITABILITY OF TRADE PROMOTIONSChannel members are then expected to "push" the product through the pipeline byoffering financial and merchandising incentives. Ultimately, the goal is to get theretailer to offer the consumer a price discount and, in many situations, merchandise theproduct through displays and newspaper advertising. In certain situations the manufac-turer will require the retailer to sign a compliance agreement which will make theretailer guarantee some form of price reduction to the consumer or merchandisingactivity on the proposed item.Most of the studies of promotional activity have concentrated on consumer responseto deals. Kuehn and Rohloff(1967) and Frank and Massy (1965) have attempted tomeasure the sales effect of price-off and other types of promotions. Webster (1965),Montgomery (1971) and Blattberget al. (1981) have attempted to measure the charac-teristics of the deal buyer. None of these studies considers how channel membersrespond to trade deals.

    Goodman and Moody (1970) developed a system to measure the effect of tradepromotions on sales of the manufacturer. Their system incorporated the buying behav-ior of other members of the channel. They postulated that the main factors affectingsales were trend and seasonality. When trade promotions occurred, the other channelmembers varied the quantity purchased, either to build inventories at lower costs orreduce inventories in anticipation of trade promotions. Their model was theoreticalwith no empirical validation.Brown (1973), using a similar concept, offered a methodology to attempt to estimatethe dealer's inventory during a deal and the trough after a deal caused by the tradeselling inventories bought during the dealing period. The problem with the methodol-ogy developed by Goodman and Moody and by Brown is that it attempts to measureconsumer purchasing indirectly. They do not attempt to estimate the size of tradeinventories, a crucial factor in determining the effect of the promotion.Abrahim and Lodish (1987) use a similar approach to Brown in estimating the effectsof alternative promotions. They use a time series model in which managers and con-sultants analyze the manufacturer's shipments in order to measure the effect of apromotion. Through different statistical procedures they create a baseline forecast ofwhat sales would be without promotions and then the difference in the baseline andactual shipments is a measure of the effect of an individual promotion. They thus alloweach promotion, even of the same type, to vary in its effect on shipments and profits.

    The major problem with Brown's and Ibrahim and Lodish's models is the possiblelack of robustness in predicting future promotional effects. By relying heavily on mod-eling residuals and allowing differing effects of identical types of promotions, it be-comes very difficult to generalize the results to future time periods. However, accordingto their published articles, they have excellent success in analyzing promotional effec-tiveness.Chevalier and Curhan (1976), using a survey, describe how trade promotions affectretailer promotion activity. While not formally modelling the relationship betweenretailer promotions and trade promotions, they show that retailers promote only alimited number of items for which they accept trade promotion allowances. Further,slow moving items are unlikely to be displayed and substantial financial inducementsare required to obtain retail price cuts. Most importantly, they question whether tradepromotions are profitable to the manufacturer.The study conducted by Chevalier and Curhan is an excellent starting point for thisarticle. Many of the issues they discuss such as "forwardbuying" (retailersloading theirinventories) and the effectiveness of differenttypes of trade promotions will be the basisof the model developed in this paper. The primary conclusions they draw will beconsistent with the findings in this paper.The model outlined in this article will combine consumer sales, factory shipments


  • 8/6/2019 Modelling Trade Promotion



    and estimated pipeline inventories to evaluate the sales effect and profitability of tradepromotion activity. In addition to developing a theoretical model, actual consumersales and shipment data are used to estimate the model's parameters and test itsaccuracy.The model described in this paper is general in form. It has been applied to differentdata sets such as SAMI as well as Nielsen bi-monthly store audit data which are used toestimate the consumer sales equation in this paper.The model can be used with scannerdata which are currently becoming available from sources such as Nielsen. As betterconsumer sales and promotional data become available, the model described in thepaper can be applied with some minor modifications to these new data sources.The remainder of the article is divided into seven sections. ?2 describes the mod