Why do large retailers discount best sellers?

You may have wondered why best sellers are often sold at considerable discounts. Sure, bestseller discounting can help retailers stimulate customer traffic and the purchase of a basket of other products in the same transaction. However, larger retailers are much more likely to do so than smaller retailers. Why is this the case?

?In our Journal of Interactive Marketing paper, Cenk, Jon and I combine game theory and Vector Autoregressive (VAR) models to answer this question. Game theory specifies the relative benefits of different scenarios for competing players, and solves for the optimal actions and game outcome assuming all players act rationally. You probably remember the ultimatum game from your studies, or Rock, Paper, Scissors from your childhood.

?Previous game theorists treated all retailer as similar (symmetric), but our paper recognized that some retailers are much larger than others, i.e. they have more products to sell. These large retailers get more benefits from attracting a customer through the discount, because these customers end up also buying other products - at healthy retail margins. This cross selling effect works both ways — customers intending to buy a best-seller may buy other items in their shopping basket, while other customers intending to buy a basket may buy a best-seller while visiting the retailer.

The pricing implications of our results reject common wisdom that loss leader pricing fails to significantly gain in-store traffic and profits. Instead, a large multi-department retailer such as Walmart can afford to turn its full Grocery business into a loss leader, making enough money on other items customers buy.

?Nice game theory, but do we have any empirical evidence? Yes, with book pricing and sales rank data from multiple online retailers. First, Amazon.com time series data shows dual causality between price and best seller rank: cheaper books sell more, but better selling books also get discounted. We specify our Vector Autoregressive (VAR) model with Discount, Rank, Number of Sellers, List price and Customer reviews as endogenous variables (explained by the model), and a constant and physical format (1 = hardcover) as exogenous variables.

The VAR results show that discounts are deeper for books with a better sales rank and better customer reviews, consistent with our game theory hypotheses. We confirmed these results in a cross-sectional analysis for a random sample of over 2 Million books across the web. Books in the top 10,000 can act as traffic generators and are discounted more heavily. Finally, we collected price and sales data on the top 100 selling books across multi-category retailers such as Walmart.com, specialized bookstores such as Borders, and book comparison sites such as bookstores.com. The top 4 retailers with highest discounts were all multi-category (45-48% discounts), while the 15 smaller retailers gave discounts between 7% (Powell's Books) and 44% (Barnes & Noble). We also demonstrate the cross-selling potential of products far down the bestseller list.

Bottom line: large retailers discount best sellers more because they are better able to monetize the resulting traffic. Is this still the case in today's environment? Let us know your thoughts!