romotion Profitability for a Retailer P the role of promotion, brand, category and store characteristics Kusum L. Ailawadi, Bari A. Harlam, et. Journal of Marketing Research Presenter: GUO DONG PROMOTION 促销 005 Contents Marketing Modeling I Promotion Profitability for a Retailer Data Methodology Correlates Key findings 1 2 3 4 03 09 13 16 DATA >> Promotion Profitability for a Retailer Data Methodology Correlates Key findings 1 2 3 4 03 09 13 16 DATA Weekly point-of-purchase scanner data for 2003 CVS ExtraCare Loyalty program, 2 years panel data Internal company records, et. CVS HOW CAN WE EVALUATE THE EFFECTS OF PROMOTION Δ SALES Δ SALES = Promotion Sales – Baseline Sales Gross Lift Manufacturer V.S. Retailer Switching Stockpiling Store net unit impact = Incremental Lift + Halo = Gross lift – Switching – Stockpiling + Halo Gross Lift = Unit sales of a item – baseline The number of weeks varies with the turnover Seasonality of the item Gross Lift Consider two extreme situations Accounting for the gross lift from all promoted items Season Store Frist difference Switching Estimate the stockpiling percentage for a category across stores Stockpiling Compare promotional buyers with nonpromotional buyers 1 2 Halo d goes from 1 to 4 for the health, beauty, edible products and general merchandise departments Separate halo effect for each department Deseasonalize and first difference Store net unit impact = Gross lift – Switching – Stockpiling + Halo Finally, net unit impact and net profit impact can be calculated: = Gross lift (1– %Switching – %Stockpiling + %Halo) Store net profit impact = Promotion sales – Base sales – Switching loss – Stockpiling loss+ Halo gain 。。。以下略