Abstract
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.
Similar content being viewed by others
Notes
Another example is Stourm et al. (2017), who study how distances between retail partners within a coalition loyalty program influence reward spillover activity.
We emphasize that we do not observe spending at competitor stores, as is typical with firm-level data. We therefore do not literally observe consolidation and business stealing effects but infer them from how changes in behavior within customers vary, especially across customers with different spatial characteristics.
See Meyer-Waarden and Benavent (2009) for an exception.
This is typical of loyalty programs in many settings, for instance an outsize share of members of airline and hotel loyalty programs are professionals who travel frequently for work.
For the average joiner, we observe about 7 months of transaction activity prior to enrollment and about 8.5 months of transactions after joining.
This loyalty program structure is consistent with that of a “customer-tier” LP, as opposed to a “frequency-based” LP, following (Blattberg, 2008).
This industry is highly concentrated, with the focal firm and the primary big-box competitor capturing about 85% of market share, according to a 2019 industry report by IBISWorld.
We selected 50 kilometers because this was the maximum radius allowed by the Google Maps API at the time of pulling this information and exceeds the average distance traveled by household by a factor of about 7.
For all tables presented in the descriptive analysis, we exclude households that fall within the upper or lower .5% of monthly sales prior to joining the loyalty program to prevent the undue influence of outliers for the summary statistics.
We selected this cutoff based on average distance driven by joining customers to surrounding focal stores.
Note that not all customers may have access to all competitors within a market. Because of this, some customers will not be associated with certain competitors.
In the web Appendices ?? and ?? we show that marketing activity is correlated with spending activity, but found no evidence that email activity is targeted based on location.
We also estimated the model with an additional four other instrumental variables, these are: proportion of customer sales in the LP category, an indicator for receiving any marketing email, the number of days between receiving the LP promotion email and the next in-category purchase, and the number of days between receiving the LP promotion email and the next purchase qualifying for LP discounts. The model with all of these instruments included in the first stage performs very similarly to the model with only the LP promotion email variable.
We show these results in the web Appendices ?? and ??
An alternative approach would be a ridge regression or similar method. In this case the penalty is applied more smoothly, shrinking coefficients on highly correlated variables towards each other. We prefer the LASSO approach because the penalty structure removes coefficients entirely, resulting in clearer model interpretation. Specifically, we can see that many non-spatial variables will drop out altogether. This result is a notable output of interest. A ridge regression would lack this clean interpretation but would result in similar quantitative predictions and can be provided upon request.
See James et al. (2013) for more information.
Due to the regularization method in LASSO, there is a tendency to bias estimates towards zero. In the post-LASSO results, the exclusion variable coefficient exhibits a slightly higher magnitude than the coefficient in the main LASSO table.
In Appendix ??, we provide additional support that the spatial variables appear to explain a substantial portion of the variation in join and change in spend activity relative to the other types of variables.
The standard errors were derived using bootstrapping. Within each targeting strategy, we sample with replacement from the observations targeted and calculate the average change in spend for that sample. The standard error is the standard deviation of this statistic across many such samples.
We would expect the predictions to be biased upwards relative to the causal treatment effect due to potential planned increases in spending from joiners. However, we can use our model to estimate the change in spend for all current non-joiners and we find that this group would increase spending by $15.58 ($1.52) per month on average (higher than the effect from actual joiners). This is evidence of adverse selection where less profitable customers are more likely to join the LP. We also consider this noteworthy because this increase is still quite low, and is substantially lower than each of the segmentation methods we consider, highlighting the valud of targeting the program.
These market types are defined in Section 3.
We attempted numerous variations on the Huff model given that it is not a direct translation of the original method, and kept only those that performed best at accurately estimating change in spending. More details are available upon request.
References
Ahn, H., & Powell, J.L. (1993). Semiparametric estimation of censored selection models with a nonparametric selection mechanism. Journal of Econometrics, 58(1-2), 3–29.
Ascarza, E. (2018). Retention futility: Targeting high-risk customers might be ineffective. Journal of Marketing Research, 55(1), 80–98.
Belloni, A., Chernozhukov, V., & et al. (2013). Least squares after model selection in high-dimensional sparse models. Bernoulli, 19(2), 521–547.
Belloni, A., Chernozhukov, V., & Hansen, C. (2014). Inference on treatment effects after selection among high-dimensional controls. The Review of Economic Studies, 81(2), 608–650.
Besanko, D., Dube, J. -P., & Gupta, S. (2003). Competitive price discrimination strategies in a vertical channel using aggregate retail data. Marketing Science, 49(9), 1121–1138.
Bijmolt, T.H., Dorotic, M., Verhoef, P.C., & et al. (2011). Loyalty programs: Generalizations on their adoption, effectiveness and design. Foundations and Trends® in Marketing, 5(4), 197–258.
Blattberg, R.C. (2008). Byung-do kim, scott a. neslin (2008), database marketing: Analyzing and managing customers. International Series in Quantitative Marketing.
Chen, Y., Li, X., & Sun, M. (2016). Competitive mobile geo targeting. Marketing Science, 36(5), 645–812.
Chen, Y., Narasimhan, C., & Zhang, Z.J. (2001). Individual marketing with imperfect targetability. Marketing Science, 20(1), 1–98.
Chen, Y., & Steckel, J.H. (2012). Modeling credit card share of wallet: Solving the incomplete information problem. Journal of Marketing Research, 49(5), 655–669.
Cook, J.A., & Siddiqui, S. (2020). Random forests and selected samples. Bulletin of Economic Research, 72(3), 272–287.
DeWulf, K., Odekerken-Schröder, G., & Iacobucci, D. (2001). Investments in consumer relationships: a cross-country and cross-industry exploration. Journal of Marketing, 65(4), 33–50.
Dreze, X., & Nunes, J.C. (2008). Feeling superior: The impact of loyalty program structure on consumers’ perceptions of status. Journal of Consumer Research, 35(6), 890–905.
Du, R.Y., Kamakura, W.A., & Mela, C.F. (2007). Size and share of customer wallet. Journal of Marketing, 71(2), 94–113.
Dube, J.-P., Fang, Z., Fong, N., & Luo, X. (2017). Competitive price targeting with smartphone coupons. Marketing Science, 36(6), 813–1017.
Ellickson, P., & Misra, S. (2012). Enriching Interactions: Incorporating Revenue and Cost Data into Static Discrete Games. Quantitative Marketing and Economics, 10, 1–26.
Fong, N.M., Fang, Z., & Luo, X. (2015). Geo-conquesting: Competitive locational targeting of mobile promotions. Journal of Marketing Research, 52(5), 726–735.
Frank, L.E., & Friedman, J.H. (1993). A statistical view of some chemometrics regression tools. Technometrics, 35(2), 109–135.
Ghose, A., & Yang, S. (2009). An empirical analysis of search engine advertising: Sponsored search in electronic markets. Management Science, 55(10), 1605–1622.
Gopalakrishnan, A., Jiang, Z., Nevskaya, Y., & Thomadsen, R. (2021). Can non-tiered customer loyalty programs be profitable? Marketing Science, 40(3), 508–526.
Hartmann, W., & Viard, B. (2008). Do frequency reward programs create switching costs? a dynamic structural analysis of demand in a reward program. Quantitative Marketing and Economics, 6(2), 109–137.
Heckman, J.J., & Robb, R. (1986). Alternative methods for solving the problem of selection bias in evaluating the impact of treatments on outcomes. In Drawing inferences from self-selected samples (pp. 63–107). Springer.
Heckman, J.J., & Robb, R. Jr (1985). Alternative methods for evaluating the impact of interventions: An overview. Journal of Econometrics, 30(1-2), 239–267.
Heckman, James. J. (1979). Sample selection bias as a specification error. Econometrica: Journal of the econometric society. 153–161.
Huff, D.L. (1964). Defining and estimating a trading area. The Journal of Marketing 34–38.
Jain, A., Misra, S., & Rudi, N. (2020). The effect of sales assistance on purchase decisions: an analysis using retail video data. Quantitative Marketing and Economics 1–31.
James, G., Witten, D., Hastie, T., & Tibshirani, R. (2013). An introduction to statistical learning Vol. 112. Berlin: Springer.
Leenheer, J., Van Heerde, H.J., Bijmolt, T.H., & Smidts, A. (2007). Do loyalty programs really enhance behavioral loyalty? an empirical analysis accounting for self-selecting members. International Journal of Research in Marketing, 24(1), 31–47.
Leung, S.F., & Yu, S (1996). On the choice between sample selection and two-part models. Journal of econometrics, 72(1-2), 197–229.
Li, Y., Gordon, B., & Netzer, O. (2018). An empirical study of national vs. local pricing by chain stores under competition. Marketing Science, 37 (5), 812–837.
Liu, Y., & Yang, R. (2009). Competing loyalty programs: Impact of market saturation, market share, and category expandability. Journal of Marketing, 73(1), 93–108.
Lockhart, R., Taylor, J., Tibshirani, R.J., & Tibshirani, R. (2014). A significance test for the lasso. Annals of statistics, 42(2), 413.
Luo, X., Andrews, M., Fang, Z., & Phang, C.W. (2014). Mobile targeting. Marketing Science, 60(7), 1738–56.
McCall, M., & Voorhees, C. (2010). The drivers of loyalty program success: An organizing framework and research agenda. Cornell Hospitality Quarterly, 51(1), 35–52.
Meyer-Waarden, L., & Benavent, C. (2009). Grocery retail loyalty program effects: self-selection or purchase behavior change? Journal of the Academy of Marketing Science, 37(3), 345–358.
Nambiar, M., Simchi-Levi, D., & Wang, H. (2019). Dynamic learning and pricing with model misspecification. Management Science, 65(11), 4980–5000.
Phillips, R., Şimşek, A.S., & Van Ryzin, G. (2015). The effectiveness of field price discretion: Empirical evidence from auto lending. Management Science, 61(8), 1741–1759.
Reilly, W. (1931). The law of retail gravitation wj reilly. New York: Inc.
Roehm, M.L., Pullins, E.B., & Roehm, H.A. Jr (2002). Designing loyalty-building programs for packaged goods brands. Journal of Marketing Research, 39(2), 202–213.
Rossi, P., McCulloch, R., & Allenby, G. (1996). The value of purchase history data in target marketing. Marketing Science, 15(4), 301–394.
Shaffer, G., & Zhang Z.J. (1995). Competitive coupon targeting. Marketing Science, 14(4), 343–466.
Shin, J., & Sudhir, K. (2010). A customer management dilemma: When is it profitable to reward one’s own customers? Marketing Science, 29(4), 671–689.
Shugan, S. (2005). Brand loyalty programs: Are they shams? Marketing Science, 24(2), 185–193.
Stourm, V., Bradlow, E., & Fader, P. (2017). Market positioning using cross-reward effects in a coalition loyalty program. HEC Paris Research Paper No. MKG-2017-1242.
Tibshirani, R. (1996). Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society: Series B (Methodological), 58 (1), 267–288.
Verhoef, P. (2003). Understanding the effect of customer relationship management efforts on customer retention and customer share development. Journal of Marketing, 67(4), 30–45.
Wachtel, S., & Otter, T. (2013). Successive sample selection and its relevance for management decisions. Marketing Science, 32(1), 170–185.
Wirtz, J., Mattila, A.S., & Oo Lwin, M. (2007). How effective are loyalty reward programs in driving share of wallet? Journal of service Research, 9(4), 327–334.
Wooldridge, J. (2002). Econometric Analysis of Cross Section and Panel Data. Cambridge: MIT Press.
Xu, X., Li, X., & Zhang, J. (2020). Regularization methods for high-dimensional sparse control function models. Journal of Statistical Planning and Inference, 206, 111–126.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix 1: Distance metrics
In the table below, x denotes the competitor type. In our empirical application this is either the big box generalist (BB), the small box generalists (SB1 or SB2), or the small box specialist (SS).
Appendix 2: Spatial variable contribution
In this supplementary section we provide an additional measure of the contribution of spatial variables to the model fit relative to other types of variables. Specifically, we compare different spatial model performance in targeting as well as highlight that our spatial metrics contribute substantially more towards R2 relative to traditional vertical variables.
1.1 Discussion of spatial model segmentation performance
In this section we discuss the relative performance of different types of spatial models at profitably segmenting LP customers. All the results discussed here correspond to Table 9. First we test a simple spatial model that takes the standard approach to spatial data of only considering distances between the focal store and customer and each of the four competitors. We see that the out-of-sample performance is actually worse than a mass marketing approach. In other words, while horizontal data is valuable in principle a simplistic approach to incorporating it that misses the nuances of competitive structure is not necessarily effective.
The second and third models extend the simple model with additional spatial metrics. First we include the count of the number of each of the four competitors within a 10 kilometer (6 miles) radius, approximately the average driving distance. Then, we augment the original spatial model with the apex angle, or the angle formed between the customer, focal store, and a given competitor with the customer at the apex. The results show that also knowing the number of nearby competitors is much more valuable than simply knowing distance between the customer and each store. In addition, knowing the angle between customer and competitors improves substantially over simply knowing distance. In terms of comparing these two data types, they perform roughly similarly as one another.
Next, to provide another comparison, we estimate three spatial models in the spirit of classic gravity models. These models date back to Reilly (1931) who coined the “law of retail gravitation” and are extended to a more flexible model by Huff (1964). These gravity models are designed to predict store choice while we have richer data on spending, but only at the focal store. We attempt to preserve the theoretical insight of these models but adjust the specification to instead predict the probability of a positive change in spend with \(p_{k=1} = \frac {w_{1}/d_{1}^{\alpha }}{{\sum }_{k=1}^{5} w_{k}/d_{k}^{\alpha }}\), where d represents the distance between the customer and each store type k (the focal store and the four competitors) and w are the weights or “attractiveness” of each store.
We consider several Huff-style model specifications. In the first, we set the store attractiveness or mass to be the square footage of each competitor type and in the second specification we treat each store attractiveness as a parameter to estimate. In both of these cases, the decay parameter α is estimated using maximum likelihood. In our third specification we allow αk to differ over store type. For segmentation, we take the models predictions and group the customers based on the percentile of positive spend in the training data.Footnote 26
We find that the gravity models perform quite well overall at identifying profitable customer groups. They outperform the mass marketing approach and the rule of thumb strategies. They also do better out-of-sample than the models using simple sales data or distance data alone, suggesting that those models are prone to overfitting in a way the more parsimonious gravity models are not. However, performance is similar to the augmented simple spatial models. In the next subsection we show that our preferred specification performs about three times as as well as the Huff-style gravity models, however.
1.2 Contribution to model fit
Table 13 presents the model fit using our two dependent variables of interest, where each row represents a single subset of the available variables. The first column displays the pseudo-R2 from a logistic regression and the second column the standard R2 from a linear regression. The variables are separated into one of five labels: our proposed spatial metrics, RFM (trip frequency and sales information), basket composition (number of items per basket, distinct items, distinct categories, etc.), the marketing indicator satisfying the exclusion restriction, and finally the selection correction polynomials from the first stage.
The spatial metrics explain considerably more than any of the alternative variable types. The sum of the R2 values across the partial models is similar to the fit from the aggregate model in the final row. This suggests that the variable types tend to be relatively orthogonal to each other. In other words, the spatial information appears to be adding a non-trivial amount of unique information to the model that would otherwise be absorbed into the residuals.
A natural rebuttal to this table is that the results are driven simply by the sheer number of spatial variables relative to the other variable types. This is only partially true: recall that all of the spatial metrics are derived solely from the latitude and longitude of the focal store, competition, and the customer. From a managerial perspective it is very easy to recreate the diverse set of spatial metrics once these few location points are obtained. In this sense, it is more the variety among the spatial variables we designed, rather than simply the number of variables, that contributes a substantial amount of information to each model.
Rights and permissions
About this article
Cite this article
Taylor, W., Hollenbeck, B. Leveraging loyalty programs using competitor based targeting. Quant Mark Econ 19, 417–455 (2021). https://doi.org/10.1007/s11129-021-09237-y
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11129-021-09237-y