Starting in the airline industry in the 1980s, loyalty programs have been adopted in businesses such as hotels, casinos, retailers, grocery stores, and restaurants. In novel research, Marketing Professor Wayne Taylor of SMU Cox and co-author Brett Hollenbeck analyze the competitive “spatial” landscape surrounding the customer and identify actionable marketing insights. Their findings suggest that the physical relationship or mapping out the customer helps in predicting profitability. But you need to know where to look.
Loyalty programs are typically structured toward rewards, discounts and managed for strategic decisions, such as targeting to increase customer engagement. Their effectiveness at increasing profits has been subject to debate, owing to the costs of administering the program itself and the benefits. Loyalty programs have the potential to increase profits by increasing switching costs for existing customers, stealing business from rivals, or through tiered discounts after a certain level of spending, or behavior-based pricing.
This paper is one of the first to explicitly link the competitive structure of a market with the performance of a loyalty program. According to Taylor, “Earlier papers have linked simple distance metrics to firm performance like how far away the customer is, to how near is the closest competitor. The key to our paper is that we really try to describe that competitive landscape in as much detail as possible.”
Predicting the Loyal
The study showed that across all customers the overall change in profits associated with customers joining the loyalty program is small or even negative. Wide variation across customers in profitability exists, as many low-spenders greatly increase their spending when they join but many of the best customers simply gain discounts on their purchases. However, certain types of program joiners seem especially important—those identified as segments of customers who seem to consolidate their purchases at the focal retailer by increasing trip frequency and customers who upgrade to higher priced products.
The authors used a large, detailed dataset on customer shopping behavior at a Fortune 500 specialty U.S. retailer before and after joining a loyalty program. The retailer specializes in product categories such as lumber, electrical, paint, and others. The loyalty program takes a common form such as tiered discounts after a certain level of spending.
The predictor of whether a loyal program customer is associated with higher or lower profits is whether they are located near a competitor store or live in an isolated market with only the focal store present. The authors say all the profit gains associated with the loyalty program are from customers in competitive markets. While intuitive, this has not been shown previously in a loyalty program setting. This motivated Taylor and Hollenbeck to use spatial data that accounts for the local competitive structure around each customer as the basis of a segmentation strategy to increase program effectiveness.
“We look at the actual physical spatial relationship between each customer, four competitors and the study’s focal store,” says Taylor. “If a focal firm is positioned in the direction of a competitor store, that actually benefits the focal firm once someone joins the loyalty program. It’s much more valuable than having someone drive in a totally opposite direction.” The study’s approach is much more complicated compared to looking at the distance between a store and a customer, which is more commonly considered. “That’s what we’re trying to capture in our research model,” Taylor offers. “If you expand how you define this spatial or physical relationship, you begin to understand more about the customer and get value out of them.”
The Spaces Around
Importantly, firms can choose which customers to invest in after they join, for example, by offering specials to a select group. Identifying the right type of customer can boost profitability. Location data adds substantial predictive power to a segmentation strategy versus historical sales patterns alone. Taylor suggests, “You just need to know the latitude and longitude of customers and that of your competitors, which can be found on Google maps. This has a lot of predictive power.”
Taylor says that the research takes all of the intuition of things on a map and turns it into something that can be modeled. What do we see on this map and how do we translate that into useful information? For example, how many stores are within a 5-mile radius; the angles of travel; stores’ density; and how saturated competitors are in a locale? “Anything relative to the consumer, the store and the competition is what we are trying to encapsulate,” he notes. Think of it as the physicality of everything.
The implications for managers of a loyalty program are to pay attention to the spatial relationships of customers. Managers focused on historical sales data for their segmentation purposes are missing out on valuable information that’s relatively easy to obtain. “We were actually quite surprised at how well these variables improved the model,” says Taylor.
“You can essentially double the profitability of targeting decisions if you also take into account spatial information—how my customers are responding to outside firms, where they are near the competition, and my potential to gain from outside firms.”
Loyalty programs are often discounted, Taylor says, but maybe managers are not looking in the right places. “People come from various places and are interacting with the world in different ways,” he concludes. “There is considerable opportunity to capitalize on customers that join a loyalty program if you know where to look.”
The paper “Leveraging Loyalty Programs Using Competitor Based Targeting” by Wayne Taylor, Cox School of Business, Southern Methodist University, and Brett Hollenbeck, University of California Los Angeles, is under review.
Written by Jennifer Warren.