Developing a Product Recommendation Model Using Spatial Statistics and Joint Space Mapping

EXTENDED ABSTRACT - A product recommendation model (Ansari, Essegaier, and Kohli 2000) is a key tool in customer relationship management (CRM). An effective recommendation model contributes to the CRM goal of customer expansion by offering high-valued products to promising regular customers. When regular customers purchase multiple products from the same company, long-term customer retention is likely to be stronger because of customers’ increased benefits and high switching costs.



Citation:

Sangkil Moon and Gary J. Russell (2005) ,"Developing a Product Recommendation Model Using Spatial Statistics and Joint Space Mapping", in AP - Asia Pacific Advances in Consumer Research Volume 6, eds. Yong-Uon Ha and Youjae Yi, Duluth, MN : Association for Consumer Research, Pages: 194-195.

Asia Pacific Advances in Consumer Research Volume 6, 2005      Pages 194-195

DEVELOPING A PRODUCT RECOMMENDATION MODEL USING SPATIAL STATISTICS AND JOINT SPACE MAPPING

Sangkil Moon, North Carolina State University, U.S.A.

Gary J. Russell, University of Iowa, U.S.A.

EXTENDED ABSTRACT -

A product recommendation model (Ansari, Essegaier, and Kohli 2000) is a key tool in customer relationship management (CRM). An effective recommendation model contributes to the CRM goal of customer expansion by offering high-valued products to promising regular customers. When regular customers purchase multiple products from the same company, long-term customer retention is likely to be stronger because of customers’ increased benefits and high switching costs.

This research develops a product recommendation model based upon the principle that customer preference similarities stemming from prior response behavior is a key element in predicting their current product purchase. The proposed product recommendation model is dependent upon two complementary methodologies: joint space mapping (placing customers and products on the same pick-any scaling map) and spatial choice modeling (specifically, the autologistic model (Besag 1974)). Briefly stated, spatial choice models (Bronnenberg and Mahajan 2001; Bronnenberg and Sismeiro 2002) assume that choices made by one customer are positively correlated with choices made by other customers with similar preferences. Using a joint space map based upon past purchase behavior, a predictive model is calibrated in which the probability of product purchase depends upon the customer’s relative distance to other customers. This approach allows the analyst to predict a given customers’ purchase probability using the model in which he purchase probability depends upon purchase behavior of similar customers. In this paper, the proposed product recommendation model considers only a single target product, reducing the model objective to finding most promising customers who will buy the given target product. An empirical analysis of this single-product recommendation model compares the proposed autologistic (AL) model along with three reasonable benchmark modelsB(1) the principal components (PC) logit model, (2) Moore-Winer (MW) model (Moore and Winer 1987), and (3) collaborative filtering (CF) (Herlocker, Konstan, and Riedl 1999). In the analysis, the autologistic model provides excellent predictions relative to the three benchmark models for a customer database of an insurance firm.

This study is intended to provide a database marketer with a new tool for identifying promising customers. From a methodological perspective, we demonstrate how ideas from spatial statistics can be used to develop a model that links customer similarity to choice behavior. In particular, the approach promises to be very effective in capturing the effects of variables that drive choice behavior, but are not explicitly included in the customer database available to the researcher. For example, variables such as lifestyle, psychographics, financial resources, and product features often determine choice behavior, but are typically absent from a firm’s database. As long as a psychometric map can be created in which locational proximity is related to product preference, the suggested methodology allows the analyst to construct a model that implicitly contains more information about customer behavior than is apparent from the available data.

The current model is limited to predict customer purchases of only a single target product. In future research, it would be worthwhile to generalize the single-product model to construct a multiple-product recommendation model that allows the researcher to integrate product similarity to deal with a multiple-product recommendation case. Eventually, exploring a further generalization of the extended multiple-product model is suggested for future research to incorporate temporal dynamics, particularly for repeatedly purchased products. In brief, the final model will be able to deal with three dimensions of purchasesB(1) customer similarity, (2) product similarity, and (3) temporal dynamics.

REFERENCES

Ansari, Asim, Skander Essegaier, and Rajeev Kohli (2000), "Internet Recommendation Systems," Journal of Marketing Research, 37 (August), 363-75.

Ariely, Dan, John G. Lynch Jr., and Manuel Aparicio IV (in press), "Learning by Collaborative and Individual-Based Recommendation Agents," Journal of Consumer Psychology.

Balabanovic, Marko and Yoav Shoham (1997), "Fab: Content-Based, Collaborative Recommendation," Communications of the Association for Computing Machinery, 40 (3), 66-72.

Besag, Julian (1974), "Spatial Interaction and the Statistical Analysis of Lattice System," Journal of the Royal Statistical Society, Series B (Methodological), 36, 192-236.

Bronnenberg, Bart J. and Vijay Mahajan (2001), "Unobserved Retailer Behavior in Multimarket Data: Joint Spatial Dependence in Market Shares and Promotion Variables," Marketing Science, 20 (Summer), 265-83.

Bronnenberg, Bart J. and Catarina Sismeiro (2002), "Using Multi-Market Data to Predict Brand Performance in Markets for Which No or Poor Data Exist," Journal of Marketing Research, 39 (February), 1-17.

Cowan, Robin, William Cowan, and Peter Swann (1997), "A Model of Demand with Interactions among Consumers," International Journal of Industrial Organization, 15, 711-732.

Cox, D. R. (1972), "The Analysis of Multivariate Binary Data," Applied Statistics (Journal of the Royal Statistical Society, Series C), 21 (2), 113-20.

Cressie, Noel A. C. (1993), Statistics For Spatial Data, Wiley Series in Probability and Mathematical Statistics. John Wiley & Sons, Inc.

Duesenberry, J. S. (1949), Income, Saving and the Theory of Consumer Behavior, Harvard University Press, Cambridge, MA.

Gershoff, Andrew D. and Patricia M. West (1998), "Using a Community of Knowledge to Build Intelligent Agents," Marketing Letters, 9 (1), 79-91.

Gruca, Thomas S., D. Sudharshan, and K. Ravi Kumar (2002), "Sibling Brands, Multiple Objectives and Responses to Entry: The Case of the Marion Retail Coffee Market," Journal of the Academy of Marketing Science, 39 (1), 59-69.

Herlocker, J., J. Konstan, and J. Riedl (1999), "An Algorithmic Framework for Performing Collaborative Filtering," SIGIR ’99: Proceedings of the 22nd Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, 230-37.

Holbrook, Morris B., William L. Moore, and Russell S. Winer (1982), "Constructing Joint Spaces from Pick-Any Data: A New Tool for Customer Analysis," Journal of Customer Research," 9 (June), 99-105.

Iacobucci, Dawn, Phipps Arabie, and Anand Bodapati (2000), "Recommendation Agents on the Internet," Journal of Interactive Marketing, 14 (3), 2-11.

Kamakura, Wagner A., Sridhar N. Ramaswami, and Rajendra K. Srivastava (1991), "Applying Latent Trait Analysis in the Evaluation of Prospects for Cross-Selling of Financial Services," International Journal of Research in Marketing, 8, 329-49.

Kamakura, Wagner A., Michel Wedel, Fernando de Rosa, and Jose Afonso Mazzon (2003), "Cross-Selling through Database Marketing: A Mixed Data Factor Analyzer for Data Augmentation and Prediction," International Journal of Research in Marketing, 20, 45-65.

Kapteyn, Arie, Sara Van De Geer, Huib Van De Stadt, and Tom Wansbeek (1997), "Interdependent Preferences: An Econometric Analysis," Journal of Applied Econometrics, 12, 665-86.

Leibenstein, H. (1950), "Bandwagon, Snob, and Veblen Effects in the Theory of Consumer’s Demand," Quarterly Journal of Economics, 64, 183-207.

Lele, Subhash (1991), "Jackknifing Linear Estimating Equations: Asymptotic Theory and Applications in Stochastic Processes," Journal of the Royal Statistical Society B, 53 (1), 253-67.

Levine, Joel H. (1979), "Joint-Space Analysis of "Pick-Any" Data: Analysis of Choices From an Unconstrained Set of Alternatives," Psychometrika, 44 (1), 85-92.

Melville, Prem, Raymond J. Mooney, and Ramadass Nagarajan (2002), "Content-Boosted Collaborative Filtering for Improved Recommendations", Proceedings of the Eighteenth National Conference on Artificial Intelligence (American Association for Artificial Intelligence).

Mooney, R. J. and L. Roy (2000), "Content-Based Book Recommending Using Learning for Text Categorization, Proceedings of the Fifth ACM Conference on Digital Libraries, 195-204.

Moore, William and Russell S. Winer (1987), "A Panel-Data Based Method For Merging Joint Space And Market Response Function Estimation, Marketing Science, 6 (Winter), 25-47.

Neslin, Scott (2002), "Churn Modeling Tournament," Teradata Center for Customer Relationship Management at Duke University.

Pazzani, Michael J. (1999), "A Framework for Collaborative, Content-Based ad Demographic Filtering," Artificial Intelligence Review, 13, 393-408.

Pollak, Robert A. (1976), "Interdependent Preferences," The American Economic Review, 66 (June), 309-20.

Russell, Gary J. and Ann Petersen (2000), "Analysis of Cross Category Dependence in Market Basket Selection," Journal of Retailing, 76 (3), 367-92.

Schafer, J. Ben, Joseph A. Konstan, and John Riedl (2001), "E-Commerce Recommendation Applications," Data Mining and Knowledge Discovery, 5, 115-53.

Solomon, Michael R. and Bruce Buchanan (1991), "A Role-Theoretic Approach to Product Symbolism: Mapping a Consumption Constellation," Journal of Business Research, 22, 95-109.

Yang, Sha and Greg M. Allenby (2003), "Modeling Interdependent Consumer Preferences," Journal of Marketing Research, Vol. 40 (August), 282-294.

----------------------------------------

Authors

Sangkil Moon, North Carolina State University, U.S.A.
Gary J. Russell, University of Iowa, U.S.A.



Volume

AP - Asia Pacific Advances in Consumer Research Volume 6 | 2005



Share Proceeding

Featured papers

See More

Featured

What Converts Webpage Visits into Crowdfunding Contributions: Assessing the Role of Circumstantial Information

Lucia Salmonson Guimarães Barros, Universidade Federal de Sao Paulo
César Zucco Jr, Brazilian School of Public and Business Administration, Brazil
Eduardo B. Andrade, FGV / EBAPE
Marcelo Salhab Brogliato, Brazilian School of Public and Business Administration, Brazil

Read More

Featured

A6. “Alexa, let’s make a trade”: Search Behavior, Trust, and Privacy with Voice-Activated Assistants

Weizi Liu, University of Illinois at Urbana-Champaign, USA
David William Ross, University of Illinois at Urbana-Champaign, USA
Kieshana M. Williams-Beeler, University of Illinois at Urbana-Champaign, USA
Yoonah Lee, University of Illinois at Urbana-Champaign, USA
Michelle Renee Nelson, University of Illinois at Urbana-Champaign, USA

Read More

Featured

When Disadvantage Is an Advantage: Benevolent Partiality in Consumer Donations

Gabriele Paolacci, Erasmus University Rotterdam, The Netherlands
Gizem Yalcin, Erasmus University Rotterdam, The Netherlands

Read More

Engage with Us

Becoming an Association for Consumer Research member is simple. Membership in ACR is relatively inexpensive, but brings significant benefits to its members.