Search results

As online reviews have become a potent marketing tool, the underlying motivation has come into question. Focus has shifted towards assessing reviewer credibility before appraising online review credibility. Guided by source credibility theory, this study investigates the effect of reviewers’ historical ratings on review helpfulness to gain insight into the role of reviewer credibility.

We explore readers’ underlying psychological processes using web data analysis (Study 1) and experiments (Study 2). Study 1 empirically examines the effect of reviewers’ historical ratings on review helpfulness using 100,621 reviews authored by 890 TripAdvisor reviewers. Study 2 involves two experiments with 328 participants to scrutinize the readers’ underlying mechanisms in establishing reviewer credibility, with a specific focus on the effect of reviewers’ historical ratings.

When a reviewer’s historical ratings are predominantly extreme, readers perceive the reviewer as less credible, leading to decreased helpfulness in reviews authored by that reviewer. Interestingly, high negativity in historical ratings does not have a significant effect on either reviewer credibility or review helpfulness.

This study offers two significant contributions to the existing literature. First, it extends previous research on review helpfulness by incorporating reviewers’ historical rating behavior. This provides a more comprehensive understanding of the factors that influence the perceived usefulness of reviews. Second, it integrates two distinct research methods: TripAdvisor web data analysis and experiments. This methodological synthesis enhances the robustness of the study by offering a more nuanced and well-rounded perspective on the dynamics between reviewers’ historical ratings and perceived helpfulness of reviews.

The aim of this research is to theorize and demonstrate that analyzing consumers’ text product reviews using text mining can enhance the explanatory power of a product sales model, particularly for hedonic products, which tend to generate emotional and subjective product evaluations. Previous research in this area has been more focused on utilitarian products.

Our text clustering-based procedure segments text reviews into multiple clusters in association with consumers’ numeric ratings to address consumer heterogeneity in taste preferences and quality valuations and the J-distribution of numeric product ratings. This approach is novel in terms of combining text clustering with numeric product ratings to address consumers’ subjective product evaluations.

Using the movie industry as our empirical application, we find that our approach of making use of product text reviews can improve the explanatory power and predictive validity of the box-office sales model.

Marketing scholars have actively investigated the impact of consumers’ online product reviews on product sales, primarily focusing on consumers’ numeric product ratings. Recently, studies have also examined user-generated content. Similarly, this study looks into users’ textual product reviews to explain product sales. It remains to be seen how generalizable our empirical results are beyond our movie application.

Whereas numeric ratings can indicate how much viewers liked products, consumers’ reviews can convey why viewers liked or disliked them. Therefore, our review analysis can help marketers understand what factors make new products succeed or fail.

Primarily our approach is suitable to products subjectively evaluated, mostly, hedonic products. In doing so, we consider consumer heterogeneity contained in reviews through our review clusters based on their divergent impacts on sales.

This study aims to explore the application of ChatGPT to analyze hotel guest satisfaction from online reviews. As online feedback plays a critical role in consumer decision-making in the hospitality industry, the research evaluates the accuracy and reliability of ChatGPT’s ratings compared to those of human raters and classic supervised machine learning classification techniques.

Using TripAdvisor reviews of five-star hotels, the authors use a structured two-phase study to assess both inter- and intra-rater reliability.

The results highlight distinct differences in rating behavior between artificial intelligence (AI) and human judges, with ChatGPT showing a tendency toward more moderate ratings. In addition, the authors observe a slight tendency for guests to overrate their experiences, supporting the literature on the subjective nature of online reviews. Despite these variations, ChatGPT shows significant agreement with guest ratings, especially when minor discrepancies are accounted for, suggesting its utility as a feedback analysis tool in the hospitality industry. This paper highlights ChatGPT’s ability to process and evaluate textual data and discusses the implications of using AI to improve review analysis processes in hospitality management. The authors advocate the incorporation of AI tools into customer feedback systems to augment human analysis and suggest future research to refine AI models for practical applications.

This study advances the understanding of AI’s role in hospitality management by demonstrating the practical application of ChatGPT for analyzing guest satisfaction through online reviews and providing a methodological framework for assessing the reliability of AI-generated content.

本研究探讨了ChatGPT在分析在线评论中的酒店顾客满意度方面的应用。由于在线反馈在酒店业消费者决策中起着关键作用, 我们的研究评估了ChatGPT评级的准确性和可靠性, 并将其与人类评价者及经典监督式机器学习分类技术进行比较。

本研究使用TripAdvisor上的五星级酒店评论, 采用结构化的两阶段研究方法, 评估评分者之间（inter-rater）和评分者自身（intra-rater）的一致性。

研究结果显示, 人工智能与人类评分者在评级行为上存在明显差异, ChatGPT表现出更倾向于中等评分的趋势。此外, 我们观察到顾客有轻微的评分偏高趋势, 这与在线评论的主观性相关的文献一致。尽管存在这些差异, ChatGPT在总体上与顾客评分具有显著一致性, 尤其是在考虑小幅评分差异时。这表明ChatGPT可作为酒店业反馈分析的有效工具。本研究强调了ChatGPT在处理和评估文本数据方面的能力, 并探讨了利用人工智能改进酒店业评论分析流程的影响。我们主张将人工智能工具纳入客户反馈系统, 以增强人工分析, 并建议未来研究进一步优化人工智能模型的实际应用。

本研究通过展示ChatGPT在分析在线评论中的顾客满意度方面的实际应用, 推动了对人工智能在酒店管理领域作用的理解, 并提供了一种评估人工智能生成内容可靠性的方法框架。

Consumers’ evaluation of online review helpfulness has been widely examined. The extant literature suggests that the attributes of review content (e.g. review length and extremity) influence review helpfulness. However, review length cannot fully reflect the richness of the review content. Anchoring on information diagnosticity and extremity bias, this study aims to explore the effect of review comprehensiveness on its helpfulness.

Field observations were obtained from 11,812 online restaurant reviews on a popular restaurant review platform. A controlled experiment was conducted to further delineate the effect of review comprehensiveness.

Review comprehensiveness moderates the effects of review length and an extremely negative review on helpfulness. It also confirms that for reviews of the same length, one covering more aspects is perceived by consumers as more helpful.

Different aspects of information in a review can efficiently assist decision-making. The results suggest that review platforms can better design their interface by providing separate areas for different product aspects. The platform can then receive more comprehensive and helpful reviews and increase the diagnosticity of these.

The study enriches the literature by introducing review comprehensiveness and examining the joint effects of review length and comprehensiveness on helpfulness. It also contributes to the literature by indicating how to reduce the effect of review extremity.

Online consumer reviews are increasingly used by third-party e-commerce organizations to shed light on the positive and negative sides of the brands they sell. However, the large number of consumer reviews requires these organizations to shortlist the most helpful ones to cope with information overload. A growing number of scholars have been investigating the determinants of review helpfulness; however, little is known about the influence of cultural factors in consumer's evaluation of review helpfulness.

This study has adopted Hofstede's cultural values framework to assess the influence of cultural factors on review helpfulness. We used a sample of 570,669 reviews of 851 hotels published by reviewers from 81 countries on Booking.com.

Findings reveal that reviewers from cultural contexts that score high on power distance, individualism, masculinity, uncertainty avoidance and indulgence are more likely to write helpful reviews.

This is one of the first cross-cultural studies in marketing using a big data approach in examining how users of reviews from different countries evaluate the helpfulness of online reviews.

Online production and transmission of child abuse material (CAM) is a complex and growing global problem. The exponential increase in the volume of CyberTips of CAM offending is placing information processing and decision-making strains on law enforcement. This paper presents the outcomes of a project that reviewed an existing risk assessment tool and then developed a new tool for CAM triaging and investigative prioritisation.

Using a mixed method approach, the authors first explored the capacity of an existing risk assessment tool for predicting a police action. The authors then used these findings to design and implement a replacement CAM decision support tool. Using a random sample of CyberTip alert cases from 2018, the authors then tested the efficiency of the new tool.

The existing risk assessment tool was not fit for CAM triaging purposes. Just six questions from the old tool were found to be statistically and significantly associated with law enforcement agents achieving a police action. The authors found that an immediate threat of abuse/endangering a child, potential case solvability, CAM image assessment, chat assessment, criticality and some weighting for professional judgement were significant in being associated with a police action. The new decision support tool is more efficient to complete and achieved a 93.6% convergence of risk ratings with the old tool using 2018 case data.

This research is unique in its development of an evidence-based decision support tool that enhances the ability of law enforcement agents to objectively and efficiently triage and prioritise increasing numbers of CyberTip alerts.

This study aims to assess the relative importance of the argument strength, argument sidedness, writing quality, number of arguments, rated review usefulness, summary review rating and number of reviews in determining the perceived usefulness and credibility of an online review. Additionally, the authors use insights from the elaboration likelihood model (ELM) to explore the effect of consumers' product category involvement on the cues' relative importance.

A conjoint analysis (N = 287) is used to study the relative importance of the seven previously mentioned attributes. A balanced orthogonal design generated eight cards that correspond to individual reviews. Respondents scored all eight cards in a random order for perceived usefulness and credibility.

Overall, argument strength is the most important cue, while summary review rating and the number of reviews are the least important for perceived review usefulness and credibility. The number of arguments is more important for people who are more highly involved with the product, while writing quality and rated review usefulness are relatively more important for the low-involvement group.

This study provides a comprehensive test of how consumers perceive online reviews, as it the first to the authors’ knowledge to simultaneously investigate a large set of cues using conjoint analysis. This method allows for the implicit valuation (utility) of the individual cues, revealing the cues' relative importance, in a setting that comes close to a real-life context. Besides, insights of the ELM are used to understand how the relative importance of cues differs depending on the level of review readers' product category involvement.

The prevalence of online review websites and the ever-growing difficulty of judging review quality result in the increasing need for consumers to reduce cognitive costs. Thus, the purpose of this study is to find out the determinants of review helpfulness based on a comprehensive theoretical framework and empirical model.

This study applied a comprehensive framework, including both review content quality and reviewer background, to investigate the determinants of review helpfulness. It also presents empirical models to further control factors around product features.

Consumers are more likely to give helpful votes to those informative and readable reviews accompanied by extreme ratings. Reviewers who disclose information, have a high reputation and report a poor experience are always identified as helpful. Consumers also tend to signal suggestions from users with a local cultural background as subjective and useless.

This study focuses on upscale hotels in China. Information registered on TripAdvisor was used presenting a residential address not nationality. Only few controlling factors available because of the limited information are shown on online review websites.

Managers of both hotels and online review websites need to focus on reviews and/or reviewers as KOLs who attract consumers’ attention and affect their subsequent decisions. A dialogue with those KOLs can be by focusing on responding to reviews with certain characteristics. A reward system for reviews and KOLs may benefit review quality on online review websites and reduce cognition costs.

This positivistic research design, with multilevel approach, presenting a comprehensive conceptual framework and empirical model not only considering review- and reviewer-related factors but also controlled factors in product or service level (hotel-related characteristics).

This paper seeks to propose and empirically validate a conceptual model on the antecedents of review helpfulness comprising three constructs, namely, valence dissimilarity, lexical dissimilarity and review order.

A panel dataset of customer reviews was collected from Amazon. Using deep learning and text processing techniques, 650,995 reviews on 13,612 products from 570,870 reviewers were analyzed. Using negative binomial regression, four hypotheses were tested.

The results indicate that new reviews with high valence dissimilarity and lexical dissimilarity compared to existing reviews are less helpful. However, over the sequence of reviews, the negative effect of review dissimilarity on review helpfulness can be moderated. This moderation differs for valence and lexical dissimilarity.

This study explains review dissimilarity in the context of online review helpfulness. It draws on the elaboration likelihood model and explains how the impacts of peripheral and central cues are moderated over the sequence of reviews.

The findings of this study provide benefits to online retailers planning to implement online reviews to improve user experience.

This paper highlights the importance of review dissimilarity in identifying user perception of online review helpfulness and understanding the dynamics of this perception over the sequence of reviews, which can lead to improved marketing strategies.

Based on more than 2.7 million online reviews (ORs) collected with big data analytical techniques from Booking.com and TripAdvisor.com, this paper aims to explore if and to what extent environmental discourse embedded in ORs has an impact on electronic word-of-mouth (e-WOM) helpfulness across eight major destination cities in North America and Europe.

This study gathered, by means of Big Data techniques, 2.7 million ORs hosted on Booking.com and TripAdvisor, and covering hospitality services in eight different destinations cities in North America (New York City, Miami, Orlando and Las Vegas) and Europe (Barcelona, London, Paris and Rome) over the period 2017–2018. The ORs were analysed by means of ad hoc content analytic dictionaries to identify the presence and depth of the environmental discourse included in each OR. A negative binomial regression analysis was used to measure the impact of the presence/depth of online environmental discourse in ORs on e-WOM helpfulness.

The findings indicate that the environmental discourse presence and depth influence positively e-WOM helpfulness. More specifically those travelers who write explicitly about environmental topics in their ORs are more likely to produce ORs that are voted as helpful by other consumers.

Implications highlight that both hotel managers and platform developers/managers should become increasingly aware of the importance that customer attach to environmental practices and initiatives and therefore engage more assiduously in environmental initiatives, if their objective is to improve online review helpfulness for other customers reading the focal reviews. Future studies might include more destinations and other operationalizations of environmental discourse.

This study constitutes the first attempt to capture how the presence and depth of hospitality services consumers’ environmental discourse influence e-WOM helpfulness on multiple digital platforms, by means of a big data analysis on a large sample of online reviews across multiple countries and destinations. As such it makes a relevant contribution to the area at the intersection between big data analytics, e-WOM and sustainable tourism research.