Accounting for Discrepancies Between Online and Offline Product Evaluations

Abstract

Despite the growth of online retail, the majority of products are still sold offline, and the "touch-and-feel" aspect of physically examining a product before purchase remains important to many consumers. In this paper, we demonstrate that large discrepancies can exist between how consumers evaluate products when examining them "live" versus based on online descriptions, even for a relatively familiar product (messenger bags) and for utilitarian features. Therefore, the use of online evaluations in market research may result in inaccurate predictions and potentially suboptimal decisions by the firm. Because eliciting preferences by conducting large-scale offline market research is costly, we propose fusing data from a large online study with data from a smaller set of participants who complete both an online and an offline study. We demonstrate our approach using conjoint studies on two sets of participants. The group who completed both online and offline studies allows us to calibrate the relationship between online and offline partworths. To obtain reliable parameter estimates, we propose two statistical methods: a hierarchical Bayesian approach and a k-nearest-neighbors approach. We demonstrate that the proposed approach achieves better out-of-sample predictive performance on individual choices (up to 25% improvement), as well as aggregate market shares (up to 33% improvement). History: K. Sudhir served as the editor-in-chief and Scott Neslin served as associate editor for this article.

The Strategic Implications of Scale in Choice-Based Conjoint Analysis

“…We chose image realism because the product-development literature suggests visual depictions and animations provide nearly the same results as physical prototypes and that rich visual representations are more realistic than text and more likely to evoke marketplace-like responses from respondents (e.g., Vriens et al 1998, Dahan and Srinivasan 2000, Dahan and Hauser 2002. Furthermore, Dzyabura et al (2019) suggest that conjoint analysis with physical prototypes provides different conjoint-analysis estimates than less realistic stimuli. Our review of the Marketing Science literature (Section 7.1) suggests that realistic images and incentive alignment are rare in the academic literature and in practice.…”

Information Acquisition and Its Incentives in an E-Commerce Supply Chain under the Offline Showroom Model

“…To illustrate Proposition 1 intuitively, we used Matlab-R2017a to draw Figure 2 based on α = 10 and θ = 0.8. Proposition 1 (1) shows that information acquisition always brings positive forecast revenue to supply chain members, which is also illustrated in Figure 2a. It is noteworthy that when commission rate is low, forecast revenue of the e-tailers is higher than that of the showroom.…”

“…To illustrate Proposition 5 intuitively, we use Matlab-R2017a to draw Figure 3 based on α = 10, θ = 0.8 and ηδ = 0.15. Comparing Theorem 2 and Proposition 5 (1), SN contract only changes the optimal information acquisition strategy. Proposition 5 (2) indicates that when information acquisition value of the showroom is relatively high (F s + S s > 2(−F i − S i )), side payment increases as bargaining power of the e-tailers increases, which is also illustrated in Figure 3.…”

“…Namely, consumers cannot make sure whether the product fits their needs before receiving the product. This results in the fact that some product categories, which need to be touched and felt, such as household products, apparel and accessories, are confronted with large amounts of returns [1,2].…”

Marketing ecosystem: An outside-in view for sustainable advantage

“…Fortunately, automated procedures for gathering unstructured data are becoming more accessible. For example, supervised machine learning methods such as artificial neural networks (Juan, Hsu, & Xie, 2017;Timoshenko & Hauser, 2019), k-Nearest Neighbors (Dzyabura, Jagabathula, & Muller, 2019), naive Bayes (Hartmann et al, 2019), and random forests (Hoornaert, Ballings, Malthouse, & Van den Poel, 2017) inductively classify textual input based on observed patterns without the requirement of manually coded rules from the researcher (Dumais, Platt, Heckerman, & Sahami, 1998). Image classification approaches that have recently made their way into marketer's toolboxes include the use of openly available computer vision packages such as Google's Cloud Vision API, Microsoft's Computer Vision API, OpenCV, Amazon's Rekognition, and IBM's Watson Visual Recognition (Mazloom, Rietveld, Rudinac, Worring, & Van Dolen, 2016).…”