Knowledge that Transforms

As new technologies diffuse within an industry, they impact the value of worker skills, and in turn may influence the mobility of workers between firms in that industry. This mobility may be particularly pronounced for individuals whose jobs are complementary to these new technologies and therefore increase demand for these individuals. In this paper, we investigate the impact of a small set of development tools diffusing within an industry and investigate whether this was associated with an increase in worker mobility for individuals with complementary skills, compared to those with (partly) substitutable ones. Specifically, we study the diffusion of middleware tools within the video game industry that made human capital more general by making it industry-specific (rather than firm-specific) and was complementary to the tasks being performed by Creatives in this industry, compared to Programmers. We exploit the uneven diffusion of these tools across genres to investigate whether workers experienced greater mobility in the genre where the technologies diffused more broadly. We also find greater mobility between projects that used the same middleware components. We find that the diffusion of these tools was associated with greater mobility for creatives than for programmers. These results are robust to a variety of empirical checks and speak to the impact of broad diffusion of IT tools, which may enable workers to more easily move between companies.

The Management Discussion and Analysis (MD&A) section in Form 10-K offers highly valuable insight of a company’s fiscal health, operational performance, and future outlook. In today’s ever-changing business environment, both practitioners and academics have extensively studied this textual data, with particular interest in quantifying the evolving information encoded in MD&A narratives. In this work, we challenge the traditional cosine similarity-based method for measuring differences between a firm’s year-over-year MD&A disclosures. We propose a novel method, D3, short for Deep Learning Method for Disclosure Differences, that incorporates external expert evaluation signals, in the form of analyst recommendation updates, to guide the learning of variant information between consecutive MD&A reports. Instead of producing a dissimilarity score as in traditional cosine distance approaches, D3 learns the changing information directly as a variant vector. The variant vector derived by D3 serves as a multifunctional artifact for various downstream financial applications. In our experiments, we show that the variant vector demonstrates strong economic utility. Specifically, a long-short portfolio strategy that sorts firms based on the magnitude of their variant vectors generates significant excess returns, after controlling for common risk factors. Moreover, the variant vector exhibits predictive power for financial risk-related outcomes. We further introduce an interpretability technique based on D3 to uncover key evolving information disclosed in MD&A reports. Our work has important implications for information systems computational design research and the emerging FinTech literature.

Non-financial information, especially information carried in disclosure reports, plays an important role in conveying financial distress signals. Considering the rise of generative AI (GenAI) and its potential in capturing both surface and latent meanings of disclosure reports, we initiate a new research avenue, GenAI-enhanced financial distress prediction. We position GenAI as an information intermediary and propose a functional analogy framework to conceptualize the process of leveraging disclosure reports with four functions: perception, extraction, reasoning, and evaluation. We then provide a guideline with three GenAI use strategies (i.e., prompt engineering, knowledge injection, and fine-tuning) and design a deep learning method featuring a function-based bidirectional representation module, which explicitly and separately extracts representations for the emphasis information produced by the extraction function and insight information produced by the reasoning function, guided by tailored convergent and divergent mutual information criteria, respectively. Empirical evaluation at the model level and impact analysis at the application level demonstrate advantages of the proposed method over benchmarked state-of-the-art methods on all fronts. Mechanism-level analyses further reveal the core drivers underlying the utility of the proposed method.

While extant studies have investigated how digital industries influence the digital transformation of traditional firms, the impact of learning from the digitalization of traditional industries has been largely overlooked. This study addresses this gap by introducing the concept of digital transformation centrality (DTC) – a novel measure of the informational advantages and associated costs a firm accrues based on its position within a network of traditional firms interconnected through their shared digital suppliers. DTC captures a new type of IT spillover: digital suppliers accumulate digital transformation knowledge by serving traditional firms, and such knowledge then spills over to other traditional firms they serve. Using extensive longitudinal data on global supply chain relationships and firm fundamentals, we find that DTC has an inverted U-shaped relationship with firm value. This shows a non-monotonic impact of digital transformation learning from traditional industries. Moreover, we develop a curvilinear mediation model to uncover the underlying mechanisms, finding that the effect of DTC is partially mediated by both innovation output and the real value of innovations. These results demonstrate the essential roles of digital suppliers in both offering digital technologies and enabling the dissemination of digital transformation knowledge among traditional firms, and provide new insights for understanding digital transformation and IT spillover across industries.

Information diffusion in social networks is uneven: some content spreads much more than other content, shaping what people see. The mix of what gets shared can leave users with a misleading sense of how often things happen. Prior research primarily examines content attributes and user attributes but has largely overlooked the role of exposure frequency—how often a user encounters an event category relative to others in their information stream. We argue that exposure frequency is a key factor influencing sharing behavior. Drawing on perceptual bias and variety-seeking, we theorize that users are more likely to share low-exposure frequency (rare) event categories. As these individual decisions accumulate, rare categories become disproportionately represented—a systematic distortion that we call rareness-biased diffusion (RBD). Across six experiments and a network simulation, we show that individuals disproportionately share rare events. At the individual level, the tendency to share rare events weakens when perceptual bias or variety seeking is suppressed but strengthens when sharing opportunities increase. Temporal clustering of rare events further reduces sharing by making rare events seem common. At the network level, distortion amplifies with distance from the source and is most stable in chain networks, while outcomes in small-world and preferential-attachment networks show greater variability due to overlapping exposure. Together, these findings introduce category-level exposure frequency as a distinct predictor of sharing, establish RBD as a new diffusion construct, and highlight implications for platform design, where simple aggregation can amplify rare events and distort public understanding.

We examine the methodological problem in which information systems (IS) research that describes itself as interpretive does not interpret meaning – in particular, the subjective meanings with which people come to understand not only the technologies they are using and managing, but also the overall lifeworld in which they live and work – with the result that meaning-related phenomena go undetected, uninvestigated, and therefore untheorized. We provide a review of how IS scholars originally emphasized meaning when interpretive research was first introduced to the IS discipline and we provide some motivating ideas pertaining to meaning that come from the phenomenology of Martin Heidegger and the phenomenology of Alfred Schutz. We cite Orlikowski (1993) as an exemplar with which to illustrate the difference that the interpretation of meaning makes to theory, as well as to point out the problem that would result if meaning were not interpreted or even acknowledged. We then examine how a study published by Califf et al. (2020) in MIS Quarterly can be considered to neglect meaning and we give additional examples of published articles that describe themselves as interpretive but are not transparent in reporting the details of any interpretation of meaning. We offer paths to correcting the methodological problem by returning to, and restoring the importance of, the basics of meaning.

Understanding the role of gender in business interactions is of significant importance to companies and society. Inspired by practical concerns of how agent gender might affect customer service interactions, the present study investigates this question using a unique dataset consisting of all public customer service interactions handled by Southwest Airlines’ Twitter account from March 2018 to September 2019. Leveraging the online text-based customer service setting where an agent’s first name serves as the only gender cue, we are able to identify the gender effect on customer behaviors and service outcomes. The identification relies on two unique features of the research context: the assignment of a customer to the next available agent is independent of agent gender; and there is no significant variation, especially gender-induced differences, in an agent’s first response. Both assumptions are supported by the data. Empirical analyses reveal that customers are more likely to continue interactions with female agents than with male agents, yet are more negative in valence of their second tweet towards female agents. Mediation tests further show that customer gender bias in their second tweets leads to downstream effects on service outcomes: interactions with female agents tend to be longer but result in lower resolution rates. Moreover, the treatment effects are moderated by customer personality and public visibility. These results offer valuable lessons to practitioners and academics regarding the unique role of gender in online customer service.

Organizations increasingly rely on social media bots for real-time monitoring. Yet, configuring bots for effective information retrieval remains challenging. Too much data creates noise; too little risks missing insights. We address this tradeoff by examining how action triggers—the search terms bots use—shape retrieval outcomes. We introduce volume-adjusted relevance, which weights relevance against retrieved volume and explore three design dimensions: semiotic specificity (hashtags vs. no-hashtags), semantic specificity (hypernyms vs. hyponyms), and trigger expansion (single vs. paired terms). In a large-scale randomized field experiment on X, a custom-built master bot retrieved over 8 million posts using 204 triggers across 50 objectives for one week. Results show that hashtags improve volume-adjusted relevance, semantic specificity provides limited benefit, and combining semantically related hashtags yields the best performance. These findings advance understanding of bot-based retrieval and offer a framework for reducing noise, avoiding blind spots, and enhancing social media monitoring.

Open source software (OSS) is an important component in the development of modern technologies, such as cloud computing and artificial intelligence algorithms. The OSS development process is characterized by diverse voluntary contributors and open membership to build software in a decentralized manner. As a result, variation in programming styles emerges in the software codebase. In this study, we introduce the concept of programming style inconsistency and argue for its importance in OSS development. As a characteristic of the codebase, programming style inconsistency arises naturally from diverse contributions and fluid membership in OSS, but it presents a challenge to effective artifact-centric coordination. We propose that programming style inconsistency, which is reflected in the differences in programming styles within software components (i.e., the component level) and across the entire codebase of the software system (i.e., the system level), is negatively associated with the technical success of OSS. Furthermore, such negative associations are moderated by two artifact-centric coordination mechanisms; namely, mitigated at the system level by modularity (coordination through codebase) and strengthened at the component and system levels by open superposition (coordination through production patterns). We test our research model with digital trace data from 1,817 JavaScript OSS projects on GitHub and measures of programming style inconsistency in the source code. Our results support the proposed hypotheses. Our study contributes to OSS coordination research by demonstrating the importance of programming style inconsistency in shaping coordination and OSS development success.