1 November 2019 in Management Science
Rapid improvements in the precision of mobile technologies now make it possible for advertisers to go beyond real-time static location and contextual information on consumers. In this paper we propose a novel “trajectory-based” targeting strategy for mobile recommendation that leverages detailed information on consumers’ physical-movement trajectories using fine-grained behavioral information from different mobility dimensions. To analyze the effectiveness of this new strategy, we designed a large-scale randomized field experiment in a large shopping mall that involved 83,370 unique user responses for a 14-day period in June 2014. We found that trajectory-based mobile targeting can, as compared with other baselines, lead to higher redemption probability, faster redemption behavior, and higher transaction amounts. It can also facilitate higher revenues for the focal store as well as the overall shopping mall. Moreover, the effect of trajectory-based targeting comes not only from improvements in the efficiency of customers’ current shopping processes but also from its ability to nudge customers toward changing their future shopping patterns and, thereby, generate additional revenues. Finally, we found significant heterogeneity in the impact of trajectory-based targeting. It is especially effective in influencing high-income consumers. Interestingly, however, it becomes less effective in boosting the revenues of the shopping mall during the weekends and for those shoppers who like to explore across products categories. Our overall findings suggest that highly targeted mobile promotions can have the inadvertent impact of reducing impulse-purchasing behavior by customers who are in an exploratory shopping stage. On a broader note, our work can be viewed as a first step toward the study of large-scale, fine-grained digital traces of individual physical behavior and how they can be used to predict—and market according to—individuals’ anticipated future behavior. This paper was accepted by Anandhi Bharadwaj, information systems.
1 June 2018 in Information Systems Research
While the growth of the mobile apps market has created significant market opportunities and economic incentives for mobile app developers to innovate, it has also inevitably invited other developers to create rip-offs. Practitioners and developers of original apps claim that copycats steal the original app’s idea and potential demand, and have called for app platforms to take action against such copycats. Surprisingly, however, there has been little rigorous research analyzing whether and how copycats affect an original app’s demand. The primary deterrent to such research is the lack of an objective way to identify whether an app is a copycat or an original. Using a combination of machine learning techniques such as natural language processing, latent semantic analysis, network-based clustering, and image analysis, we propose a method to identify apps as original or copycat and detect two types of copycats: deceptive and nondeceptive. Based on the detection results, we conduct an econometric analysis to determine the impact of copycat apps on the demand for the original apps on a sample of 10,100 action game apps by 5,141 developers that were released in the iOS App Store over five years. Our results indicate that the effect of a specific copycat on an original app’s demand is determined by the quality and level of deceptiveness of the copycat. High-quality nondeceptive copycats negatively affect demand for the originals. By contrast, low-quality, deceptive copycats positively affect demand for the originals. Results indicate that in aggregate the impact of copycats on the demand of original mobile apps is statistically insignificant. Our study contributes to the growing literature on mobile app consumption by presenting a method to identify copycats and providing evidence of the impact of copycats on an original app’s demand. The online appendix is available at https://doi.org/10.1287/isre.2017.0735 .
1 March 2019 in Management Science
It is now well understood that social media plays an increasingly important role in consumers’ decision making. However, an overload of social media content in product search engines can hinder consumers from efficiently seeking information. We propose a structural econometric model to understand consumers’ preferences and costs on search engines to improve user experience under unstructured social media. Our model combines an optimal stopping framework with an individual-level random utility choice model and analyzes click behavior in conjunction with purchase choices. Our model accounts for three major constraints in a consumer’s decision-making process: (1) interdependency in decision making for different alternatives, (2) sequential arrival of information revealed by click-throughs, and (3) nonnegligible search cost. Our approach allows us to jointly estimate consumers’ heterogeneous preferences and search costs under the interplay of social media and search engines, and to predict search and purchase behavior for each consumer. We validate the model using an individual session-level data set of approximately seven million observations resulting in room bookings in 2,117 U.S. hotels. Interestingly, our analysis allows us to quantify the trade-off between consumers’ benefits and cognitive costs from using large-scale unstructured social media information during decision making. Our policy experiments show that providing a carefully curated digest of social media content during the earlier stages of consumer search (i.e., on the search results summary page) can lead to a 12.01% increase in the overall search engine revenue. The online appendix is available at https://doi.org/10.1287/mnsc.2017.2991 . This paper was accepted by Anandhi Bharadwaj, information systems.
1 September 2019 in Information Systems Research
Low engagement rates and high attrition rates have been formidable challenges to mobile apps and their long-term success. To date, little is known about how companies can scientifically detect user engagement stages and optimize corresponding personalized-targeting promotion strategies to improve business revenues. This paper proposes a new structural forward-looking hidden Markov model as combined with a randomized field experiment on app notification promotions. Our model can recover consumer latent engagement stages by accounting for both the time-varying nature of users’ engagement and their forward-looking consumption behavior. The structural estimates from the FHMM with the field-experimental data enable us to identify heterogeneity in the treatment effects. Additionally, we simulate and optimize the revenues of different personalized-targeting promotion strategies with the structural estimates. Personalized dynamic engagement-based targeting based on the FHMM can generate substantially higher revenues than the experience-based targeting strategy applied by current industry practices and targeting strategies based on alternative customer segmentation models. Overall, the novel feature of our paper is its proposal of a new personalized-targeting approach combining the FHMM with a field experiment to tackle the challenge of low engagement with mobile apps.
1 September 2020 in Information Systems Research
Consumers rely on review platforms when deciding where to stay, where to eat, what movies to watch, or even which doctor to use. This is great for consumers, but it has makes online review platforms a target for fraud. Review platforms have responded by developing tools and algorithms to identify potentially fraudulent reviews. But the question remains: What should platforms do with fraudulent reviews after detecting them? Our research answers this question using randomized experiments and large-scale data analysis from Yelp’s review platform. Our results show that after detecting fraudulent reviews, platforms should keep them on their platforms, but should display them with a flag that identifies them as potentially fraudulent. Doing so will increase consumers' trust in the platform by demonstrating that the platform takes fraud serious and will also penalize dishonest businesses. Together, these results provide strong managerial and policy guidance to developing truthful, transparent, and accountable online ecosystems. Our research topic is particularly timely given the presence of misinformation on technology platforms, the incentives of actors to exploit anonymity to manipulate consumer beliefs, and the influence these actions can have on consumer trust.
1 March 2020 in Information Systems Research
Banks today have been increasingly reducing their physical presence and redirecting customers to digital channels, and yet, the consequences of this strategy are not well studied. This research investigates the effects of banks’ branch network changes (i.e., branch openings and branch closures) on customer omnichannel banking behavior. Using a proprietary data set from a large commercial bank in the United States, this paper shows the asymmetric effects of branch openings and branch closures on customer omnichannel banking behavior. In particular, it finds that branch openings increase customers’ branch transactions. However, the first branch opening leads to a migration of complex transactions to the branches, which might result in a net decrease in online banking in the short term. As consumers interact more with the physical channel, there is a gradual synergistic increase in customers’ transactions via online banking as well as alternative channels due to a learning spillover effect. This learning spillover effect goes from easy online inquiries to more complex online transactions as additional branches open. On the contrary, branch closures result in a favorable migration pattern from the branch channel to online banking. This pattern, however, could be reversed once the last branch closes within the customer’s residential neighborhood.