12 Feb 2007Working Paper Summaries

Adding Bricks to Clicks: The Effects of Store Openings on Sales through Direct Channels

by Jill Avery, Mary Caravella, John Deighton and Thomas Steenburgh

Consider a retailer who operates both brick-and-mortar stores and direct channels such as direct mail catalogs and an Internet Web site. What effect does the opening of a new retail store have on direct channel sales in the retail trading area surrounding the store? Does the existence of more opportunities for consumer contact with the brand increase the retailer's direct sales, or does intra-brand, inter-channel competition erode the retailer's direct sales? Does consumer response to the retailer's brand evolve over time, perhaps as consumers go through some process of trial-and-error learning about the relative merits of stores and direct channels, or is the impact of the new store relatively discrete? Does the answer depend on whether consumers in the retail trading area have had the opportunity for previous experience with the brand's stores? This research used a proprietary longitudinal dataset from a multichannel retailer to understand what happens and to probe the implications for channel management strategy. Key concepts include:

Adding a physical retail store to existing direct sales channels increases firm sales in the long run, as sales from the new store are incremental to sales from direct channels, which show little long term damage from channel competition.
Adding channels produces both cannibalizing and complementary effects which operate in tandem and vary over time. Cannibalization occurs in the short term following the addition of a new channel, while complementarity takes time to manifest itself. Retail store openings cannibalize direct channel sales in the short term if physical stores do not already exist in the retail trading area, but produce complementary effects which overcome the losses from cannibalization in the long run.
Our results suggest the underlying consumer shopping behavior driving this result. The opening of a retail store may induce some existing direct channel customers to switch their purchases to the retail store; simultaneously, new customers are attracted to the direct channels, perhaps due to a branding effect stemming from the publicity surrounding the new store which makes customers more aware of and more comfortable with the firm's direct channel operations.
Use caution extrapolating these results to other retailers. This study involved only store openings by a single retailer with a well established and respected brand into markets where the retailer did not previously have stores. Direct retailers with less established brands may benefit even more than this retailer from branding effects by opening a new store.

Author Abstract

We assess the effect of opening physical retail stores on direct channel sales. Our data come from a leading U.S. retailer which opened four new stores; two openings occurred in retail trading areas which had been previously served by direct channels alone and two openings occurred in retail trading areas which had been served by both direct channels and existing physical stores. We hypothesize two effects, cannibalization and complementarity, and conjecture that the magnitude of these effects may change over time. We find that retail store openings cannibalize direct channel sales in the short term if physical stores do not already exist in the retail trading area, but produce complementary effects which overcome the losses from cannibalization in the long run. Our results are based on both interrupted time series analysis and difference-in-differences analysis of six years of proprietary sales data and we use a novel zip code matching method to better isolate the effects of channel expansion. We argue for advantages to using zip code level data for methodological and consumer data privacy reasons.