Asim Ansari

Internet Recommendation Systems

Several online firms, including Yahoo!, Amazon.com, and Movie Critic, recommend documents and products to consumers. Typically, the recommendations are based on content and/or collaborative filtering methods. The authors examine the merits of these methods, suggest that preference models used in marketing offer good alternatives, and describe a Bayesian preference model that allows statistical integration of five types of information useful for making recommendations: a persons expressed preferences, preferences of other consumers, expert evaluations, item characteristics, and individual characteristics. The proposed method accounts for not only preference heterogeneity across users but also unobserved product heterogeneity by introducing the interaction of unobserved product attributes with customer characteristics. The authors describe estimation by means of Markov chain Monte Carlo methods and use the model with a large data set to recommend movies either when collaborative filtering methods are viable alternatives or when no recommendations can be made by these methods.

Hierarchical Bayes Versus Finite Mixture Conjoint Analysis Models: A Comparison of Fit, Prediction, and Partworth Recovery

A study conducted by Vriens, Wedel, and Wilms (1996) and published in Journal of Marketing Research found that finite mixture (FM) conjoint models had the best overall performance of nine conjoint segmentation methods in terms of fit, prediction, and parameter recovery. Since that study, hierarchical Bayes (HB) conjoint analysis methods have been proposed to estimate individual-level partworths and have received much attention in the marketing research literature. However, no study has compared the relative effectiveness of FM and HB conjoint analysis models in terms of fit, prediction, and parameter recovery. To conduct such a comparison, the authors employ the simulation methodology proposed by Vriens, Wedel, and Wilms with some modification. The authors estimate traditional individual-level conjoint models as well. The authors show that FM and HB models are equally effective in recovering individual-level parameters and predicting ratings of holdout profiles. Two surprising findings are that (1) HB performs well even when partworths come from a mixture of distributions and (2) FM produces good parameter estimates, even at the individual level. The authors show that both models are quite robust to violations of underlying assumptions and that traditional individual-level models overfit the data.

Choice and the Internet: From Clickstream to Research Stream

The authors discuss research progress and future opportunities for modeling consumer choice on the Internet using clickstream data (the electronic records of Internet usage recorded by company web servers and syndicated data services). The authors compare the nature of Internet choice (as captured by clickstream data) with supermarket choice (as captured by UPC scanner panel data), highlighting the differences relevant to choice modelers. Though the application of choice models to clickstream data is relatively new, the authors review existing early work and provide a two-by-two categorization of the applications studied to date (delineating search versus purchase on the one hand and within-site versus across-site choices on the other). The paper offers directions for further research in these areas and discusses additional opportunities afforded by clickstream information, including personalization, data mining, automation, and customer valuation. Notwithstanding the numerous challenges associated with clickstream data research, the authors conclude that the detailed nature of the information tracked about Internet usage and e-commerce transactions presents an enormous opportunity for empirical modelers to enhance the understanding and prediction of choice behavior.

A Model of Consumer Learning for Service Quality and Usage

In many services (e.g., the wireless service industry), consumers choose a service plan according to their expected consumption. In such situations, consumers experience two forms of uncertainty. First, they may be uncertain about the quality of their service provider and can learn about it after repeated use of the service. Second, they may be uncertain about their own usage of minutes and learn about it after observing their actual consumption. The authors propose a model to capture this dual learning process while accounting for the nonlinearity of the pricing scheme used in wireless services. The results show that both quality learning and quantity learning are important. The authors conduct several policy experiments to capture the effects of consumer learning, pricing, and service quality on customer lifetime value (CLV). They find that consumer learning can result in a win–win situation for both consumers and firm; consumers leave less minutes on the table, and the firm experiences an increase in overall CLV. For example, the authors find that there is a 35% increase (approximately

Modeling Multiple Relationships in Social Networks

75) in overall CLV with consumer learning than without. The key driver of this result is the change in the retention rate with and without learning.

Bayesian factor analysis for multilevel binary observations

Firms are increasingly seeking to harness the potential of social networks for marketing purposes. Therefore, marketers are interested in understanding the antecedents and consequences of relationship formation within networks and in predicting interactivity among users. The authors develop an integrated statistical framework for simultaneously modeling the connectivity structure of multiple relationships of different types on a common set of actors. Their modeling approach incorporates several distinct facets to capture both the determinants of relationships and the structural characteristics of multiplex and sequential networks. They develop hierarchical Bayesian methods for estimation and illustrate their model with two applications: The first application uses a sequential network of communications among managers involved in new product development activities, and the second uses an online collaborative social network of musicians. The authors’ applications demonstrate the benefits of modeling multiple relations jointly for both substantive and predictive purposes. They also illustrate how information in one relationship can be leveraged to predict connectivity in another relation.

Heterogeneous factor analysis models: A bayesian approach

Multilevel covariance structure models have become increasingly popular in the psychometric literature in the past few years to account for population heterogeneity and complex study designs. We develop practical simulation based procedures for Bayesian inference of multilevel binary factor analysis models. We illustrate how Markov Chain Monte Carlo procedures such as Gibbs sampling and Metropolis-Hastings methods can be used to perform Bayesian inference, model checking and model comparison without the need for multidimensional numerical integration. We illustrate the proposed estimation methods using three simulation studies and an application involving students achievement results in different areas of mathematics.

Consumers' Price Sensitivities Across Complementary Categories

Multilevel factor analysis models are widely used in the social sciences to account for heterogeneity in mean structures. In this paper we extend previous work on multilevel models to account for general forms of heterogeneity in confirmatory factor analysis models. We specify various models of mean and covariance heterogeneity in confirmatory factor analysis and develop Markov Chain Monte Carlo (MCMC) procedures to perform Bayesian inference, model checking, and model comparison.We test our methodology using synthetic data and data from a consumption emotion study. The results from synthetic data show that our Bayesian model perform well in recovering the true parameters and selecting the appropriate model. More importantly, the results clearly illustrate the consequences of ignoring heterogeneity. Specifically, we find that ignoring heterogeneity can lead to sign reversals of the factor covariances, inflation of factor variances and underappreciation of uncertainty in parameter estimates. The results from the emotion study show that subjects vary both in means and covariances. Thus traditional psychometric methods cannot fully capture the heterogeneity in our data.

Representing Heterogeneity in Consumer Response Models 1996 Choice Conference Participants

In this paper, we examine the pattern of correlation among consumer price sensitivities for customer purchase incidence decisions across complementary product categories. We use a hierarchical Bayesian multivariate probit model to uncover this pattern. We estimated this model using purchase incidence data for six categories involving three pairs of complementary products. Our results show a new and interesting pattern of correlation among price parameters of complementary products. For example, we find that the correlation of own-price sensitivities of complementary products is negative. These results are consistent across the three complementary pairs of products. We also investigate the reason for this counterintuitive result. Finally, we present some managerial implications of our model. We show how our model can be used for cross-category targeting decisions by retailers. We find that compared to nontargeted discounting, the average profitability gain from customized discounting across the three category pairs is only 1.29% when complementarity is ignored, but this gain improves to 8.26% when full complementarity is taken into account. We also investigate whether ignoring the complex pattern of correlation has implications for managerial actions regarding targeting and optimal discounting. We find that retailers can make misleading inferences about the impact of targeted discounts when they ignore cross-category effects in modeling.

Dynamic Targeted Pricing in B2B Relationships

We define sources of heterogeneity in consumer utility functions relatedto individual differences in response tendencies, drivers of utility, formof the consumer utility function, perceptions of attributes, statedependencies, and stochasticity. A variety of alternative modelingapproaches are reviewed that accommodate subsets of these various sourcesincluding clusterwise regression, latent structure models, compounddistributions, random coefficients models, etc. We conclude by defining anumber of promising research areas in this field.