Incorporating Experience Quality Data into CRM Models: The Impact of Gambler Outcomes on Casino Return Times
Enabled by modern interaction-logging technologies, managers increasingly have access to data on
quality levels in customer interactions. We consider the direct marketing targeting problem in situations
where 1) the customer's experience quality level varies randomly and independently from occasion to
occasion, 2) the rm has measures of the quality levels experienced by each customer on each occasion,
and 3) the rm can customize marketing according to these measures and the customer's behaviors. A
primary contribution of this paper is a framework and methodology to use data on customer experience
quality data to model a customer's evolving beliefs about the rm's quality and how these beliefs combine
with marketing to influence purchase behavior. Thereby, this paper allows the manager to assess the
marketing response of a customer with any specific experience and behavior history, which in turn can be
used to decide which customers to target for marketing. This research develops a novel, tractable way to
estimate and introduce flexible heterogeneity distributions into Bayesian learning models. The model is
estimated using data from the casino industry, an industry which generates more than $60 billion in U.S.
revenues but has surprisingly little academic, econometric research. The counterfactuals offer interesting
findings on gambler learning and direct marketing responsiveness and suggest that casino profitability
can increase substantially when marketing incorporates gamblers' beliefs and past outcome sequences
into the targeting decision.
Leveraging Loyalty Programs Using Competitor Based Targeting
Loyalty programs (LPs) are widely used by firms but not well understood. These programs provide discounts and perks to loyal customers and are costly to administer, but produce uncertain changes in spending patterns. We use a large and detailed dataset on customer shopping behavior at one of the largest U.S. retailers before and after joining a loyalty program to evaluate how behavior changes. We combine this with detailed spatial data on customer and store locations, including the locations of competing firms. We find significant changes in behavior associated with joining the LP with a large amount of heterogeneity across customers. We find that location relative to competitors is the factor most strongly associated with increases in spending following joining the LP, suggesting that the LP's quantity discounts work primarily through business stealing and not through other demand expansion. We next estimate a model of what variables determine how spending will change after joining the LP. We use high-dimensional data on spatial relationships between customers, the focal firm's stores, and competing stores as well as customers' historical spending patterns. This model is used to test whether past sales data reflecting customer's vertical value to the firm or spatial data reflecting customer's horizontal vulnerability are more important determinants of post-LP spending increases. We show how LASSO regularization estimated on complex spatial relationships are more effective than are models using past sales data or simpler spatial models. Finally, we show how firms can use customer and competitor location data to substantially increase LP performance through spatially driven segmentation.
Controlling for Retailer Synergies when Evaluating Coalition Loyalty Programs: A Bayesian Additive Regression Tree Approach
Spatial models in retailing allow for correlations among purchase decisions from consumers within predefined geographic areas. The purpose of these models is to control for unobserved demand side effects at the regional-level (e.g., a neighborhood), but they typically ignore synergies among individual retailers within a region. To capture the synergies on both the supply side (e.g., store density) and the demand side (e.g., socioeconomic differences across regions), we augment a traditional spatial model with a Bayesian Additive Regression Tree (BART). This allows us to account for unobserved regional differences and observed but potentially complex interactions among individual customers and retailers. We apply this model to a credit card coalition loyalty program (CLP). In our empirical setting, we are interested in analyzing the impact of the loyalty program earnings structure on monthly spend. We do this while controlling for the evolving coalition network, which contains hundreds of geographically dispersed partner retailers. Our data has two key features that permit this. First, the retail partner network evolves over time; this variation in retailer participation allows us to observe card spending patterns when individual retailers are both in and out of the coalition network. Second, the data contains a natural experiment where the loyalty program changed its earning structure, which allows us to estimate the impact of the rewards rate of the loyalty program on customer spend. Our findings show that failure to control for the dynamics of the coalition network results in severely biased estimates of CLP rewards effectiveness. We discuss the implications of BART in our empirical setting and highlight its potential in other marketing situations which contain numerous, interacting control variables.