Flexible Decomposition of Sales Promotion Effects Using Store-Level Scanner Data

Dick Wittink (), Peter S.H. Leeflang () and Harald J. van Heerde ()
Additional contact information
Dick Wittink: School of Management
Peter S.H. Leeflang: Faculteit der Economische Wetenschappen
Harald J. van Heerde: Faculty of Economics and Business Administration

Yale School of Management Working Papers from Yale School of Management

Abstract: Recent studies in marketing show decompositions of sales promotion effects based on household-level scanner data. Typically, the total elasticity is decomposed into choice, timing, and quantity elasticities. We propose a model that estimates standard, enhanced, and flexible decompositions based on store-level data. The standard decomposition divides the own-brand sales elasticity into a market share ("choice") elasticity, a period share ("timing") elasticity, and a category sales ("quantity") elasticity. The enhanced decomposition includes two extensions. One extension is the separation of the category sales elasticity into a between-store share and a total market elasticity, which allows us to measure cross-store effects of promotions. The other extension is a partition of the market share elasticity into within-brand and between-brand share elasticities. We show that the store-level results for standard decompositions based on four product categories are comparable to extant household-level results. For the two extensions, our empirical results, for one product category each, show small between-store share elasticities but large cannibalization effects between SKUs of the same brand. For improved understanding of promotion effects, we propose flexible decompositions. One is that we allow the decomposition to depend on the type of promotion support, such as feature and/or display. We find that the own-brand sales elasticity increases with more support. Also, for more support, the relative contribution of the period share elasticity increases while the contribution of the market share elasticity decreases. The second type of flexibility is that we allow the decomposition to depend on the magnitude of the discount. We use a flexible, nonparametric, method to account for the latter dependency. We find that with display and/or feature, the own-brand sales elasticities are less negative at higher price discount levels. The decomposition shows that on average, the market share elasticity rapidly becomes less negative while the period share- and the category sales elasticities are fairly constant when the discount increases. Hence, the relative contribution of the latter two elasticities increases at the cost of the relative contribution of the market share elasticity. The flexible estimation of promotion effects, separately for four types of support, provides superior insight into marketplace phenomena relevant to marketing scientists and marketing practitioners.

Keywords: Econometric Modeling; Nonparametric Estimation; Price Elasticity; Promotion; Regression and Other Statistical Techniques (search for similar items in EconPapers)
JEL-codes: C14 M31 (search for similar items in EconPapers)
Date: 2001-05-31
References: Add references at CitEc
Citations:

Downloads: (external link)
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=271292 (application/pdf)
Our link check indicates that this URL is bad, the error code is: 410 Gone (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=271292 [301 Moved Permanently]--> https://papers.ssrn.com/sol3/papers.cfm?abstract_id=271292)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ysm:somwrk:ysm193

Access Statistics for this paper

More papers in Yale School of Management Working Papers from Yale School of Management Contact information at EDIRC.
Bibliographic data for series maintained by ().