JFJuan Feng

Juan Feng

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Juan Feng has a total of 38 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • The Divorce of Word and Deed—A Data-Mining Approach to Identify and Evaluate Customer Requirements

    by Kexin Ai, Juan Feng, Xinyu Sun

    Whereas online reviews have become a primary data source for understanding customer requirements in both research and practice, using such information alone to guide product design can be unreliable. Our research investigates whether and how consumers’ preferences expressed through online reviews (words) align with their actual purchase decisions (deeds). Our analysis shows that features praised in online reviews do not necessarily translate to market success. This inconsistency between what consumers say and what they do poses significant challenges for manufacturers in product development decisions. We empirically identify the existence of word–deed inconsistency in consumer preferences. Some features are silent in online reviews yet significantly drive purchase decisions, whereas others are frequently praised but have limited influence on actual purchases. Building on these insights, we propose an innovative dual-weights model that extends existing two-dimensional customer requirement analysis by integrating both prepurchase choice drivers and postpurchase satisfaction determinants. Using this model, we classify features based on their importance for satisfaction versus purchase decisions and offer actionable product improvement strategies for different types of features.

  • Better Is Better? Signaling Paradoxes in Performance-Based Advertising

    by Ran Pan, Juan Feng

    Under performance-based advertising, firms pay only for measurable consumer actions, such as clicks or sales. However, these actions may result from factors other than effective advertising. For example, a high-reputation seller is more likely to generate more actions than a low-reputation seller. If such consumer actions are counted toward advertising performance, a high-quality firm is penalized because it has to pay more than necessary to signal its quality. Without addressing these paradoxes, it becomes increasingly challenging for a high-quality firm to signal its quality under the performance-based advertising scheme. Our research extends advertising signaling theory, which primarily focuses on new products without an established reputation and assumes that the advertising reaches the entire market (e.g., Milgrom and Roberts 1986 , Feng and Xie 2012 ). Instead, we consider products that may have an established reputation (e.g., newer versions of the iPhone) and allow for the scenario that not every market segment can be reached by the advertising. We identify two paradoxes that arise from performance-based advertising and discuss methods to mitigate these paradoxes by proposing different ways of measuring advertising performance.


  • SaaS or On-premises? Compete Through Customizability, Price, and Hybrid Offerings

    by Nan Li, Kunhao Jia, Feng Juan

    This paper studies the competition between two software vendors: one offering software as a service (SaaS) and the other providing on-premises software, both facing software upgrades with different customizability and implementation costs. We build a theoretical model to explore the vendor's pricing and market segmentation strategies. We find that (1) enhancing the customizability of SaaS software does not necessarily benefit the SaaS vendor, nor does it necessarily harm the on-premises vendor, depending on the current level of the SaaS software's customizability; (2) The SaaS vendor generally prefers pay-per-use pricing to subscription pricing because it can not only serve more consumers but also help relieve competition pressure from the on-premises vendor; (3) Adopting a hybrid model is not necessarily profit enhancing. Specifically, while the SaaS vendor always benefits from adopting a hybrid pricing strategy, the on-premises vendor may be better off by either not offering or offering a SaaS version with limited customizability, even if it does not make sales. Nevertheless, both vendors adopting a hybrid model leads to socially efficient competition outcome, where both vendors and consumers are better off, a win-win-win scenario.


  • Open Access

    MiniFed : Integrating LLM-based Agentic-Workflow for Simulating FOMC Meeting

    by Sungil Seok, Shuide Wen, Qiyuan Yang, Juan Feng, Wenming Yang

    The Federal Funds rate in the United States plays a significant role in both domestic and international financial markets. However, research has predominantly focused on the effects of adjustments to the Federal Funds rate rather than on the decision-making process itself. Recent advancements in large language models(LLMs) offer a potential method for reconstructing the original FOMC meetings, which are responsible for setting the Federal Funds rate. In this paper, we propose a five-stage FOMC meeting simulation framework, MiniFed, which employs LLM agents to simulate real-world FOMC meeting members and optimize the FOMC structure. This framework effectively revitalizes the FOMC meeting process and facilitates projections of the Federal Funds rate. Experimental results demonstrate that our proposed MiniFed framework achieves both high accuracy in Federal Funds rate projections and behavioral alignment with the agents' real-world counterparts. Given that few studies have focused on employing LLM agents to simulate large-scale real-world conferences, our work can serve as a benchmark for future developments.

  • The Emergence of a Sharing Market: Pricing, Supply, and Consumption

    by Ling Ding, Juan Feng, Xiuwu Liao, Lu Yang

    The emerging sharing economy increases the utilization of existing products/services, reduces the consumption of new resources, and gets more popular in recent years. We consider a market where consumers are differentiated by their preference for the product and develop an analytical model to study how a person-to-person (P2P) sharing market (ideal or imperfect) affects the manufacturer’s pricing strategies, consumers’ consumption time, consumer surplus, and social welfare. Surprisingly, we find that the existence of the sharing market may drive up the selling price of the product if the sharing platform can endogenously determine the commission fee of sharing transactions. Accordingly, depending on the transaction costs and the proportion of high-type consumers, we find that consumers’ total consumption time, consumer surplus, manufacturer’s profit, and social welfare may not improve in the presence of a sharing market. This study also offers guidelines for the sharing platform and optimal pricing strategies for manufacturers.


  • Does Money Talk? The Impact of Monetary Incentives on User-Generated Content Contributions

    by Yuewen Liu, Juan Feng

    Many platforms use monetary incentives to encourage user-generated content (UGC) contributions. The empirical evidence, however, is contradictory: monetary incentives are shown to either increase or decrease contribution. We make the first attempt to build a unified theoretical model to understand the complex nature of the impact of monetary incentives. We consider contributors differentiated not only by their attitudes toward monetary incentives but also by their effectiveness to attract audience. We identify two scenarios where contributors can be crowded out when monetary incentives are present: (1) when a small amount of monetary incentive is introduced, the non–money-driven contributors reduce or even stop contributing (motivation crowding out); or (2) when the monetary incentive is relatively large, the high-effectiveness contributors crowd out the low-effectiveness ones (competition crowding out). As a result, an increase in the monetary incentive can either increase or decrease contributors’ participation and the total content volume contributed. Our results offer guidelines for different UGC platforms to design monetary incentive mechanisms.


  • Online Product Reviews-Triggered Dynamic Pricing: Theory and Evidence

    by Juan Feng, Xin Li, Xiaoquan (Michael) Zhang

    Online product reviews are arguably one of the most easily accessible sources of marketing data for online retailers. It is possible to build machine learning tools to learn consumers' opinions from online word of mouth (WOM). Menu costs are practically trivial for online retailers, and it is not difficult to program automatic price changes based on live feeds of online review data. This paper argues that sellers can use online product reviews to develop better pricing strategies. We first build a theoretical model to examine a seller's optimal pricing strategy when online WOM information is taken into consideration. We find that, with consumer reviews, firms may take price-skimming and penetration strategies depending on the combination of consumer characteristics (such as misfit cost) and product characteristics (such as product quality). We examine a book retailing data set collected from online stores to offer empirical support for the analytical predictions.


  • Open Access

    A Causal Inference Method for Reducing Gender Bias in Word Embedding Relations

    by Zekun Yang, Juan Feng

    Word embedding has become essential for natural language processing as it boosts empirical performances of various tasks. However, recent research discovers that gender bias is incorporated in neural word embeddings, and downstream tasks that rely on these biased word vectors also produce gender-biased results. While some word-embedding gender-debiasing methods have been developed, these methods mainly focus on reducing gender bias associated with gender direction and fail to reduce the gender bias presented in word embedding relations. In this paper, we design a causal and simple approach for mitigating gender bias in word vector relation by utilizing the statistical dependency between gender-definition word embeddings and gender-biased word embeddings. Our method attains state-of-the-art results on gender-debiasing tasks, lexical- and sentence-level evaluation tasks, and downstream coreference resolution tasks.

  • When Online Reviews Meet Sales Volume Information: Is More or Accurate Information Always Better?

    by Yang Liu, Juan Feng, Xiuwu Liao

    This paper studies how two types of market-generated information, namely, online reviews and past sales volume information, jointly affect consumer purchase decisions as well as firms’ pricing strategies. We build a two-period duopoly model in a market with herding consumers who have different preferences and are unsure of the quality difference between products. In addition, a firm’s sales volume is uncertain because of the existence of “irregular” consumers. We find that the impacts of online reviews and sales volume information on firms’ profits are mutually enhancing. The impact of such market-generated information depends on both product characteristics (the level of consumers’ misfit for the nonpreferred product) and how consumers a priori perceive the quality difference between the products. Contrary to the conventional wisdom that more/accurate information is beneficial to high-quality firms, as well as consumers, we find that such information can be detrimental to firms, as firms adjust prices to induce, or react to, favorable market-generated information. Accordingly, consumers may not benefit from such market-generated information if the gain from less uncertainty cannot offset the loss from higher product prices. Such findings offer guidelines for firms to design better pricing strategies, as well as information policies, such as whether or not to promote more informative reviews. The online appendix is available at https://doi.org/10.1287/isre.2017.0715 .