Search

Online reviews significantly impact consumers’ decision-making process and firms’ economic outcomes and are widely seen as crucial to the success of online markets. Firms, therefore, have a strong incentive to manipulate ratings using fake reviews. This presents a problem that academic researchers have tried to solve for over two decades and on which platforms expend a large amount of resources. Nevertheless, the prevalence of fake reviews is arguably higher than ever. To combat this, we collect a dataset of reviews for thousands of Amazon products and develop a general and highly accurate method for detecting fake reviews. A unique difference between previous datasets and ours is that we directly observe which sellers buy fake reviews. Thus, while prior research has trained models using laboratory-generated reviews or proxies for fake reviews, we are able to train a model using actual fake reviews. We show that products that buy fake reviews are highly clustered in the product reviewer network. Therefore, features constructed from this network are highly predictive of which products buy fake reviews. We show that our network-based approach is also successful at detecting fake review buyers even without ground truth data, as unsupervised clustering methods can accurately identify fake review buyers by identifying clusters of products that are closely connected in the network. While text or metadata can be manipulated to evade detection, network-based features are more costly to manipulate because these features result directly from the inherent limitations of buying reviews from online review marketplaces, making our detection approach more robust to manipulation.

Judging by the millions of reviews left by guests on the Airbnb platform, this "trusted community marketplace" is fulfilling its mission of matching travelers seeking accommodation with hosts who have room to spare remarkably well. Based on our analysis of ratings we collected for over 600,000 properties listed on Airbnb worldwide, we find that nearly 95% of Airbnb properties boast an average user-generated rating of either 4.5 or 5 stars (the maximum); virtually none have less than a 3.5 star rating. We contrast this with the ratings of approximately half a million hotels worldwide that we collected on TripAdvisor, where there is a much lower average rating of 3.8 stars, and more variance across reviews. Considering properties by accommodation type and by location, we find considerable variability in ratings, and observe that vacation rental properties on TripAdvisor have ratings most similar to ratings of Airbnb properties. Last, we consider several thousand properties that are listed on both platforms. For these cross-listed properties, we find that even though the average ratings on Airbnb and TripAdvisor are similar, proportionally more properties receive the highest ratings (4.5 stars and above) on Airbnb than on TripAdvisor. Moreover, there is only weak correlation in the ratings of individual cross-listed properties across the two platforms. Our work is a first step towards understanding and interpreting nuances of user-generated ratings in the context of the sharing economy.

Marketers commonly seed information about products and brands through individuals believed to be influential on social media, which often involves enlisting micro influencers, users who have accumulated thousands as opposed to millions of followers (i.e., other users who have subscribed to see that individual’s posts). Given an abundance of micro influencers to choose from, cues that help distinguish more versus less effective influencers on social media are of increasing interest to marketers. The authors identify one such cue: the number of users the prospective influencer is following. Using a combination of real-world data analysis and controlled lab experiments, they show that following fewer others, conditional on having a substantial number of followers, has a positive effect on a social media user’s perceived influence. Further, the authors find greater perceived influence impacts engagement with the content shared in terms of other users exhibiting more favorable attitudes toward it (i.e., likes) and a greater propensity to spread it (i.e., retweets). They identify a theoretically important mechanism underlying the effect: following fewer others conveys greater autonomy, a signal of influence in the eyes of others.

Are reviews with photos more helpful? If so, do consumers find reviews more helpful when photos and text convey similar or different information? This paper examines the effect of content similarity between photos and text on review helpfulness and its underlying mechanism. Using a dataset of 7.4M reviews associated with 3.5M photos from Yelp, and applying machine learning algorithms, we quantify the similarity of the content between text and photos. We find that, overall, photos increase the helpfulness of a review. More importantly, though, greater similarity between photos and text heightens review helpfulness more. We then validate algorithm-based similarity assessments with similarity perceptions of human judges. Using real-world reviews from Yelp and carefully designed stimuli, we replicate our core findings in five laboratory experiments. Further, testing the underlying mechanism, we find that greater similarity facilitates ease with which consumers can process the review which, in turn, increases that review’s helpfulness to consumers. Finally, we show that factors that impede processing ease (e.g., language difficulty or poor image quality) can reduce the effect of similarity on helpfulness. These findings provide novel insights into the value of user-generated content that includes text and photos and its underlying mechanism.

Isogeometric analysis (IGA) has been a particularly impactful development in the realm of Kirchhoff–Love thin-shell analysis because the high-order basis functions employed naturally satisfy the requirement of C 1 continuity. Still, engineering models of appreciable complexity, such as wind turbine blades, are typically modeled using multiple surface patches and, often, neither rotational continuity nor conforming discretization can be practically obtained at patch interfaces. A penalty approach for coupling adjacent patches is therefore presented. The proposed method imposes both displacement and rotational continuity and is applicable to either smooth or non-smooth interfaces and either matching or non-matching discretization. The penalty formulations require only a single, dimensionless penalty coefficient for both displacement and rotation coupling terms, alleviating the problem-dependent nature of the penalty parameters. Using this coupling methodology, numerous benchmark problems encapsulating a variety of analysis types, geometrical and material properties, and matching and non-matching interfaces are addressed. The coupling methodology produces consistently accurate results throughout all tests. Furthermore, the suggested penalty coefficient of α = 1 0 3 is shown to be effective for the wide range of problem configurations addressed. Finally, a realistic wind turbine blade model, consisting of 27 patches and 51 coupling interfaces and having a chordwise- and spanwise-variant composite material definition, is subjected to buckling, vibration, and nonlinear deformation analyses using the proposed approach.

Is visual-verbal word-of-mouth more persuasive? If so, do consumers find communication more helpful when photos and text convey similar or different information? This paper examines the effect of photo-text similarity on review helpfulness and its underlying drivers. Using a dataset of 6.8M reviews including 3.3M photos from Yelp and applying state-of-the-art machine learning algorithms, we quantify the similarity of the content between text and photos. We find that it is not only the mere presence of a photo that increases helpfulness but also the similarity between the photo content and the review text. We replicate our main findings and examine the underlying drivers in two laboratory experiments. When the photos and text convey similar (vs. dissimilar) information, consumers find the review more helpful because 1) the information in the review becomes easier to process, and 2) quality inferences of the focal attribute are heightened. However, greater similarity 3) limits the total amount of information conveyed, reducing helpfulness. We find that drivers (1) and (2) outweigh (3). Therefore, the totality of these three distinct processes allows greater photo-text similarity to heighten persuasion. These findings provide novel insights into the persuasiveness of visual-verbal word-of-mouth and its underlying psychological drivers.

Machine learning is bringing us self-driving cars, medical diagnoses, and language translation, but how can machine learning help marketers improve marketing decisions? Machine learning models predict extremely well, are scalable to “big data,” and are a natural fit to analyze rich media content, such as text, images, audio, and video. Examples of current marketing applications include identification of customer needs from online data, accurate prediction of consumer response to advertising, personalized pricing, and product recommendations. But without the human input and insight—the soul—the applications of machine learning are limited. To create competitive or cooperative strategies, to generate creative product designs, to be accurate for “what-if” and “but-for” applications, to devise dynamic policies, to advance knowledge, to protect consumer privacy, and avoid algorithm bias, machine learning needs a soul. The brightest future is based on the synergy of what the machine can do well and what humans do well. We provide examples and predictions for the future.

The sharing economy, in which ordinary consumers also act as sellers, is attracting much interest from scholars and practitioners. The rapid growth of this sector of the economy and the emergence of several large brands like Uber and Airbnb raise several interesting questions that need to be researched. We therefore review and summarize the extant research on this sector and suggest several additional directions for future research.

Studying competition and market structure at the product level instead of brand level can provide firms with insights on cannibalization and product line optimization. We introduce Product2Vec, a method based on representation learning, to study product-level competition when the number of products is large. The proposed model takes shopping baskets as inputs and, for every product, generates a low-dimensional vector that preserves important product information. Using these product vectors, we present several findings. First, we show that these vectors can recover analogies between product pairs. Second, we create two measures, complementarity and exchangeability, that allow us to determine whether product pairs are complements or substitutes. Third, we combine these vectors with traditional choice models to study product-level competition. To accurately estimate price elasticities, we modify the representation learning algorithm to remove the influence of price from the product vectors. We show that, compared with state-of-the-art choice models, our approach is faster and can produce more accurate demand forecasts and price elasticities. Fourth, we present two applications of Product2Vec to marketing problems: 1) analyzing intra- and inter-brand competition and 2) analyzing market structure. Overall, our results demonstrate that machine learning algorithms, such as representation learning, can be useful tools to augment and improve traditional marketing methods.

We present a computational framework for applying the phase-field approach to brittle fracture efficiently to complex shell structures. The momentum and phase-field equations are solved in a staggered scheme using isogeometric Kirchhoff–Love shell analysis for the structural part and isogeometric second- and fourth-order phase-field formulations for the brittle fracture part. For the application to complex multipatch structures, we propose penalty formulations for imposing all the required interface constraints, i.e., displacement () and rotational () continuity for the structure as well as and continuity for the phase field, where the latter is required only in the case of the fourth-order phase-field model. All involved penalty terms are scaled with the corresponding problem parameters to ensure a consistent scaling of the penalty contributions to the global system of equations. As a consequence, all coupling terms are controlled by one global penalty parameter, which can be set to independent of the problem parameters. Furthermore, we present a multistep predictor–corrector algorithm for adaptive local refinement with LR NURBS, which can accurately predict and refine the region around the crack even in cases where fracture fully develops in a single load step, such that rather coarse initial meshes can be used, which is essential especially for the application to large structures. Finally, we investigate and compare the numerical efficiency of loosely vs. strongly staggered solution schemes and of the second- vs. fourth-order phase-field models., Computer Methods in Applied Mechanics and Engineering, 372, ISSN:0045-7825, ISSN:1879-2138

In this paper, we study the effect of Airbnb on restaurant revenue in the state of Texas. Using a difference in difference strategy that exploits the geographical and temporal variation of Airbnb diffusion in Texas, we show that a 1% increase in Airbnb reviews in a zipcode is associated with a 0.011% increase in restaurant revenue in the same zipcode. Considering the median annual Airbnb growth in each zipcode, this result implies that Airbnb can explain about 12% of the median annual restaurant revenue growth. Moreover, we show that this increase in revenue is driven by an increase in restaurant demand that stems from additional tourism demand generated by Airbnb. Finally, we show that the positive effect of Airbnb is more pronounced for independent restaurants and in less commercial—and therefore with less economic activity—zipcodes. Overall, our results provide evidence that home-sharing platforms like Airbnb can increase economic activity for businesses and in neighborhoods that are more likely to need it the most.