Personalized promotions within food retail - which strategies are most profitable for retailers vs suppliers?


Within food retail its becoming increasingly common for firms to charge different prices to different customers based on their past purchase behaviour, often referred to as “personalized promotions”. The challenge for retailers and manufacturers is to provide the right customers with the right promotions at the right time. To do that you have to answer questions such as "which customers should receive personalized promotions?", "should the retailer share customer  data with manufacturers?", "which behaviour should trigger a personalized discount?" In a recent study published in Journal of Marketing Science, researchers try to answer these questions by investigating the profit impact from different personalized promotion strategies for retailers vs suppliers. The study is based on household scanner panel data on cola purchases of 356 households making 32 942 shipping trips at a US suburban supermarket store.


What we know about personalized promotions

Before we look at the results from the study, let’s review what research tells us about personalized promotions:

  • First of all, research and empirical studies suggest that the effectiveness of price promotions varies across customers. This means it will be more profitable to send promotions to some customers (with high promotion elasticity) than others (with low promotion elasticity).
  • Secondly, an extensive empirical literature suggests that customers’ promotion elasticities changes over time – and, that on account of switching costs in demand, a customer's promotion elasticity for a the most recently purchased brand decreases at his/her next purchase occasion. For example, a customer that has recently purchased a Pepsi will respond less to a price promotion for a Coca Cola.

Retailers that want to optimize the profit impact form their personalized promotions must take these factors into account. Also, being a category profit maximizer, the retailer’s pricing incentive may not align with the manufacturers’ pricing incentives. What is the optimal for the retail category manager, may not align with the desired outcome of the managers of the manufacturer brands.

In the study, the researches have simulated the profit impact from different personalized promotion strategies within the soft drink category, from the retailer’s and the manufacturers’ perspective. The study is based on scanner panel data on cola purchases of 356 households making 32942 shopping trips at a supermarket store (which is a local monopolist) in a suburban market of a large U.S. city. Here are the key findings:

  • First of all, the retailer can more than double (x2,2) the incremental profit from personalized promotions by using the most recent purchase information of each customer in addition to the customer segment membership (+11% compared to a profit of 0,9 USD when there are not personalized promotions) vs using segment membership only (+5% compared to profit when there are no promotions). This increase is available at little to no analytical cost since tracking the most recent purchase outcome of a customer is very easy from a database management standpoint for a retailer.
  • Secondly, when the retailer engages in BPD, manufacturer profit increases for the large manufacturer (Pepsi) but decreases for the small manufacturer (Coke).
  • Finally, we the retailer benefits (the retailer’s profit increases) when both manufacturers jointly employ personalized discounts (and the retailer helps them distribute them), as opposed to the retailer employing personalized discounts by itself. If the retailer can charge the manufacturers for access to its customer database, its profit improves further. However, the manufacturers end up being worse off than under the case of no personalized price promotion.

Best regards, the Formulate team

Arvid Stenback Lund