How to harness the potential of user-generated content for management decisions

Kübler, Raoul; Burmester, Alexa; Paetz, Friederike; Klarmann, Martin Article How to harness the potential of user-generated content for management decisions Schmalenbach Journal of Business Research (SBUR) Provided in Cooperation with: Schmalenbach-Gesellschaft für Betriebswirtschaft e.V. Suggested Citation: Kübler, Raoul; Burmester, Alexa; Paetz, Friederike; Klarmann, Martin (2025) : How to harness the potential of user-generated content for management decisions, Schmalenbach Journal of Business Research (SBUR), ISSN 2366-6153, Springer, Heidelberg, Vol. 77, Iss. 3, pp. 407-418, https://doi.org/10.1007/s41471-025-00226-5 This Version is available at: https://hdl.handle.net/10419/331931 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ EDITORIAL https://doi.org/10.1007/s41471-025-00226-5 Schmalenbach Journal of Business Research (2025) 77:407–418 How to Harness the Potential of User-Generated Content for Management Decisions Raoul V. Kübler · Alexa Burmester · Friederike Paetz · Martin Klarmann Accepted: 28 August 2025 / Published online: 1 September 2025 © The Author(s) 2025 1 The Evolving Concept of User-Generated Content Almost one and a half decades ago, Google’s CEO Eric Schmidt pointed out that humankind creates and stores more data every single day than was generated in the previous three thousand years combined (Schmidt 2010). While the majority of these trillions of bytes originate from non-human sources such as machine sensors (which themselves often measure and track human behavior), an ever-growing share of data is actively created by human users through online channels and digital environments. Within this data, individuals (i.e., consumers) constantly share information about which products, services, and brands they like or dislike, consider, test, or use. Although terminologies such as user-generated content (UGC), user-created content (UCC),oruser-generated data can be traced back to pre-internet eras, such as the 1980s (see, e.g., Kaplan and Haenlein 2010), the concept of UGC gained widespread attention with the rapid diffusion of Web 2.0 technologies. These technologies enabled users not only to create but also to disseminate content easily through online channels. At this point, consumers began actively sharing information about their everyday lives, including their consumption experiences. While some attempts have been made to conceptualize UGC as a form of electronic word-of-mouth (eWOM), it is important to note that UGC encompasses more than “positive or negative state- Raoul V. Kübler Marketing, ESSEC Business School, Cergy, France E-Mail: [email protected] Alexa Burmester Applied Quantitative Methods, Kühne Logistics University, Hamburg, Germany Friederike Paetz Mathematics, Statistics, and Applied Computer Science, Hochschule Anhalt, Bernburg, Germany Martin Klarmann Marketing, Karlsruhe Institute of Technology, Karlsruhe, Germany K 408 Schmalenbach Journal of Business Research (2025) 77:407–418 ments made by potential, actual, or former customers about a product or company, which is made available to a multitude of people and institutions via the Internet” (Hennig-Thurau et al. 2004, p. 39). UGC does not necessarily have to address products, brands, companies, or services; it may also reflect a consumer’s general mood, emotions, opinions, or other experiences. Kaplan and Haenlein (2010), drawing on an OECD (2007) definition, highlight three main criteria that define UGC: (1) the content is at least partly publicly accessible, (2) it reflects some degree of creative effort, and (3) it is created with nonprofessional intent. Their approach was primarily driven by the need to classify the growing body of content actively created by users on social media. However, the characteristics of UGC have since expanded through new technologies (see, e.g., Kübler and Hennig-Thurau 2025), use cases (see, e.g., Shirazi et al. 2013 demonstrating how messenger apps can be used to track user’s sleeping patterns), and the increasing opportunities to trace consumers’ digital footprints in the online environment (see, e.g., Kosinski et al. 2013). Whereas Kaplan and Haenlein (2010) emphasize the importance of creative effort, UGC in today’s world may also emerge from non-creative interactions with digital technologies. These interactions still generate valuable data that can be leveraged to predict relevant outcomes. For example, Althoff et al. (2025) use step counts recorded by smartphones to examine how a city’s architectural design influences physical activity. Similarly, search engine queries and Wikipedia access data have been employed to predict flu outbreaks or the number of COVID-19 infections (see, e.g., Skiera et al. 2020), while simple liking and following data on social media can be used to infer personal traits, values, and interests (Kosinski et al. 2013). Given the informational potential of data arising from interactions between users and digital touchpoints, we propose a broader conceptualization of user-generated content. While users may actively create content (e.g., in the form of social media posts), they can also passively generate equally information-rich data through their interactions with platforms and technologies. This data, too, can be fruitfully leveraged by both scholars and managers. We therefore suggest extending the initial OECD (2007) definition as follows: 1. Publication: The content or data are made accessible to others via the internet, either explicitly (e.g., posts, reviews, videos) or implicitly through digital platforms (e.g., likes, follows, search queries, or wearable data that platforms aggregate and display). Content can either be directly obtained from the platform through access points such as App Programming Interfaces (APIs) or through automated data extraction tools such as crawlers or scrapers. 2. Creative or Interactional Effort: While UGC traditionally reflects deliberate creative effort, it can also result from users’ purposeful engagement with digital technologies, where actions such as searching, liking, or wearing a connected device generate data traces. 3. Outside Professional Routines: UGC is typically produced outside formal professional or institutionalized processes, emerging from everyday practices of individuals as they communicate, search, share, or interact online. K Schmalenbach Journal of Business Research (2025) 77:407–418 409 Over the past decade, UGC has been increasingly explored, monitored, and utilized in business research, also giving rise to several special issues. Early work in the field focused on understanding the motives behind why and how individuals contribute UGC (see Fader and Winer 2012). More recently, attention has shifted toward examining which methods and tools can be employed to extract and refine information from UGC (see, e.g., Vomberg et al. 2024). With the growth of UGC, new tools and methods have become available to scholars, enabling them to analyze UGC and extract valuable information from it (de Haan et al. 2024). Although UGC is often characterized by its unstructured nature, advances in natural language processing as well as classification and object recognition techniques from information sciences now allow researchers to transform unstructured data—such as text, images, audio files, or video—into structured formats that can be incorporated into the traditional econometric models commonly used in business research (see, e.g., Yildirim and Kübler 2023). 2 How to Harness the Potential of User-Generated Content for Management Decisions Recent progress in generative AI, large language models (LLMs), transformer architectures, and diffusion models has made it easier than ever to extract insights from unstructured UGC—often rapidly and at low cost—by deriving variables such as customer sentiment or brand mentions (see, e.g., Feuerriegel et al. 2025; Hartmann et al. 2023; Kübler et al. 2020). The rapid advancement of such tools holds even greater potential. Emerging developments in generative AI technologies open new avenues for leveraging UGC to inform decision-making better and strengthen business strategies. While an increasing number of studies now rely on these tools and data harvested from online sources—and thus incorporate UGC—as either dependent or independent variables (see, e.g., Boegershausen et al. 2022, for a detailed overview of UGC-related research in marketing), relatively little research has systematically investigated for which domains UGC is most suitable and how managerial decision-making can effectively draw on this novel form of data, which is often freely available to both scholars and practitioners. This special issue seeks to explore the synergies of new methods and available UGC sources and to shed light on the trends, challenges, and opportunities shaping this evolving landscape. 3 Leveraging UGC for Business and Marketing Research Figure 1summarizes the steps required to understand better the potential synergies between recently developed analytical tools and the vast amount of UGC available to business scholars and managerial decision-makers. The first step is to establish a deeper understanding of the use cases and informational potential offered by the different sources of UGC. This requires a systematic K 410 Schmalenbach Journal of Business Research (2025) 77:407–418 Fig. 1 Leveraging UGC in Business and Marketing Research review, synthesis, and analysis of recent work in the field, which can then be used to highlight future application areas. Such an approach provides clear pathways for advancing research by identifying promising domains, clarifying which application cases are feasible, and outlining the types of insights and data required—both analytically and empirically—to make these applications viable. In addition, a systematic conceptual analysis of available UGC sources is necessary to understand the types of content users create, as well as how UGC both influences and is influenced by consumer decision-making processes (see e.g. Hildebrand and Schlager 2019) and which factors drive the production of UGC (see e.g., Shriver et al. 2013). Mapping these UGC sources and linking them to stages of the customer journey will allow researchers to identify the most suitable types of UGC for specific research objectives, thereby guiding future work toward more targeted and impactful applications. While business research has long benefited from established routines for measuring (often latent) consumer and customer attitudes using validated scales, UGCbased research remains in its infancy. In the case of latent attitude measurement, clear procedures ensuring construct reliability and validity have been developed and refined for more than half a century (see, e.g., Peter 1979). By contrast, UGC research still resembles a “wild west,” where scholars frequently operationalize conK Schmalenbach Journal of Business Research (2025) 77:407–418 411 structs (e.g., sentiment) using the cheapest or simplest tools available, often without sufficient attention to construct validation or the potential reliability issues associated with their approach (Berger et al. 2020). As a third priority, it is thus important to develop clear guidelines on how to transfer the knowledge gained by decades of psychometric research into UGC operationalization and how to develop valid and reliable UGC measures. With the wide availability of UGC—and the increasing accessibility of powerful generative AI tools for its analysis—ethical considerations have become more pressing than ever. Large volumes of data can now be easily leveraged to infer sensitive personal information. On the one hand, this can help to foster consumer well-being (see e.g., Dewender and Kübler 2025) and help vulnerable entities (see e.g., Skiera et al. 2020; Sukhwal et al. 2023). On the other hand, UGC may also be used in ways that conflict with users’ own interests (see, e.g., Kosinski et al. 2013;O’Neil 2017;Jainetal.2016), or be leveraged to manipulate public opinions (Pauwels et al. 2025), distribute disinformation (Vosoughi et al. 2018), or affect political decision making (Kübler et al. 2025b). It is therefore imperative not only to highlight the potential of UGC but also to establish clear ethical guidelines for researchers. This is not merely a matter of moral responsibility; it is also essential for preserving consumer trust. Once eroded, such trust may ultimately undermine the availability of UGC itself, as consumers could become reluctant to continue sharing information online—whether actively or passively. While the digital UGC world appears to offer virtually limitless resources for research, the offline world is increasingly marked by dwindling resources. Although concerns about the environmental footprint of generative AI and LLMs—such as high energy consumption—are valid, a similar pressing challenge lies in understanding how sustainability itself is communicated and perceived. UGC provides a unique lens into how individuals and organizations express, interpret, and respond to sustainability-related topics. Exploring these signals can help identify gaps between corporate messaging and consumer experience, detect greenwashing risks, and ultimately support more transparent, accountable, and effective sustainability strategies. As such, UGC is not only a tool for advancing managerial insight but also a critical resource for shaping and evaluating sustainable business practices. With the five papers included in this special issue, we attempt to address each of the five points highlighted in the above framework. 4 The Content of this Special Issue On the Potential of User-Generated Content for Management Decisions In the following, we discuss how the content of this special issue ties to the framework displayed in Fig. 1, and then present in Table 1an overview of the contributions made by each article. K 412 Schmalenbach Journal of Business Research (2025) 77:407–418 Table 1 Overview of articles included in this special issue Authors Issue Response Contribution Baier et al. (2025) How can firms collect and analyze UGC to support marketing decisions across areas like market selection, brand, quality, and NPD? Reviews UGC collection (web crawlers/scrapers, APIs) and analysis methods (feature extraction, traditional ML, discriminative & generative deep learning), plus representative use cases. Provides a practical guideline for selecting data sources and methods Comprehensive overview of methods & use cases; adapts a structured threestep approach to guide decision-making with UGC; flags ethics/privacy and representativeness issues; sketches future research avenues Schröder et al. (2025) What role does UGC play along the customer journey? Which stages and drivers are most studied, and what should managers do? Structured review of 342 articles; develops a framework mapping UGC effects across pre-purchase, purchase, post-purchase and six dimensions (UGC, product, writer, consumer, interaction, other). Finds research concentrates on pre-purchase and UGC characteristics; contrasts reviews vs. social media patterns; offers managerial actions (encourage creation, monitor, respond) Integrates dispersed findings into a journey-stage× driver framework; identifies under-researched stages/ contexts; delivers actionable guidance for UGC management and a research agenda Kübler et al. (2025a) Can UGC approximate consumer mindset metrics (e.g., awareness, consideration, intention, satisfaction) along the decision journey, overcoming survey limits? Argues for leveraging UGC to approximate survey-based CMMs and proposes a four-step process: (1) identify construct facets, (2) identify data sources, (3) choose transformation/ML tool, (4) (implied) track/validate Introduces a structured path to mine/ track latent mindset metrics from UGC at scale; highlights constructvalidity considerations and practical alignment with the customer decision journey Christ et al. (2025) What does digital responsibility mean for UGD-based decision-making? How can UGD be governed responsibly across its lifecycle (capture!use) and across direct (core) and indirect (peripheral) stakeholders? Who is responsible, for what, and to whom? Proposes a Responsible UGD Governance Model integrating three dimensions—object (UGD capture/use; data-as-facts vs. data-as-world-making), authority (core vs. peripheral ecosystems; hard/soft law), and subject ((joint) controllership)—and emphasizes transparency, contextual integrity, and distributed accountability Articulates governance principles and actionable checkpoints; connects to evolving regulation (GDPR, AI Act, DSA, Data Act); and lays out a research agenda for responsible UGD governance Blits et al. (2025) How do firms and users communicate sustainability (triple bottom line) in FGC and UGC, and how can it be analyzed? Proposes a conceptual signaling model and a custom sustainability dictionary; demonstrates an illustrative pipeline across corporate websites, Amazon reviews, and YouTube videos to detect environmental, social, and economic signals Supplies tools (dictionary+ workflow) for TBL content mining across FGC/UGC; surfaces research questions and practical guidance for monitoring/managing sustainability communication K Schmalenbach Journal of Business Research (2025) 77:407–418 413 5 UGC Selection and Analysis While many research studies have relied on UGC as either an independent or dependent variable, a systematic investigation of its use cases—as well as a synthesis of the facilitators and challenges associated with working with UGC—is still lacking (Guyt et al. 2024). Although research widely acknowledges the value of UGC for information-enriched decision-making (Berger et al. 2020; Boegershausen et al. 2022), a clear categorization of current and potential use cases remains absent. At the same time, insights into the validity of UGC as a proxy for various businessrelated measures are essential in order to integrate secondary, UGC-based metrics with traditional, primary data approaches (de Haan et al. 2024). Addressing this need requires a systematic discussion of the diverse collection and analysis tools available to UGC scholars. The first article of this special issue—Baier et al. (2025)—tackles this gap by providing a comprehensive overview of suitable methods and use cases of UGC in business and marketing research. They deliver an end-to-end playbook for UGCbased decision support—covering how to collect data (crawlers, scrapers, APIs), transform it, and analyze text, image, audio, and video—illustrated with use cases across market selection, brand, quality, and NPD. On the analytics side, they organize the toolkit from feature-engineering+ ML to discriminative deep learning (Convolutional Neural Networks (CNNs)/Transformers) and LLM-enabled generative workflows (summarization, topic discovery), clarifying when each is most suitable. They also distill practical guidance (e.g., transfer learning, multimodal pipelines) and flag key constraints. The authors ultimately propose a structured three-step approach that assists researchers in identifying relevant UGC sources, selecting appropriate methods for feature extraction, and transforming unstructured data into structured formats. In addition, they highlight critical knowledge gaps, such as the representativeness of UGC data in the context of generating consumer insights and the need for a better understanding of which UGC sources and categories tie to which phases of the customer journey. 6 UGC Sources Along the Decision-Making Journey The second article of this special issue—Schröder et al. (2025)—subsequently addresses this gap by investigating the role UGC plays along the customer journey. They code 342 top-tier marketing articles (2014–2024) into a 3-stage journey (pre-, purchase, post-purchase) and a six-driver taxonomy: (1) UGC characteristics, (2) customer characteristics, (3) product characteristics, (4) writer characteristics, (5) interaction characteristics, and (6) other. This reveals a strong concentration of evidence in the pre-purchase stage (e.g., intentions, attitudes, helpfulness) and around UGC characteristics themselves, while the purchase and post-purchase stages and several contexts remain comparatively underexplored. The framework’s two axes (stage and driver) make it easy to locate prior findings and generate hypotheses. Along the stage axis, dependent variables typical of each phase are used to anchor results (e.g., pre-purchase: intentions/attitudes/ K 414 Schmalenbach Journal of Business Research (2025) 77:407–418 helpfulness; purchase: sales/choice; post-purchase: usage/loyalty). Practically, the framework translates into where to act and what to test. For managers, it underlines three near-term moves: (i) stimulate creation (e.g., incentives, on-site prompts), (ii) monitor systematically (social listening across reviews and social feeds), and (iii) respond deliberately (especially to negative content)—because UGC demonstrably shapes outcomes at all stages. For researchers, the grid spotlights under-researched cells, notably purchase/post-purchase outcomes, utilitarian/credence goods, user interactions around reviews, and designs that compare UGC types side-by-side rather than in isolation. It also flags a fast-moving frontier: how AI-generated content might alter UGC dynamics along the journey. 7 UGC Measure Validity and Reliability While the first and second articles of this special issue highlight the widespread use of UGC in business research, they also make clear that much of the existing work does not adequately examine whether UGC-derived measures truly provide valid and reliable instruments, and thus whether they accurately capture common consumer attitudes such as satisfaction, consideration, or other latent constructs. As in survey-based research, it is imperative that UGC-based measures are both free from error—yielding consistent results—and valid in the sense that they genuinely measure the constructs they are intended to capture (see Peter 1979 for a detailed discussion of scale reliability and validity). The third article in this special issue—Kübler et al. (2025a)—offers a first framework to transfer insights from psychometric research into the development of UGCbased measures. Focusing on customer mindset metrics (CMMs)—the most widely applied form of psychometrics in marketing—the paper positions UGC as an alwayson measurement layer and formalizes a psychometrics-anchored four-step process for constructing such measures. The framework proceeds by deriving construct facets from established scales and distinguishing brandversus product-level targets; selecting UGC sources that can express these facets given access, volume, and legal constraints; identifying appropriate extraction tools suited to the data format and cadence; and, finally, validating the measures with a strong emphasis on predictive validity along the customer decision journey to ensure managerial usefulness. Beyond this framework, the article provides actionable guidance by mapping which UGC sources are most appropriate for different CMMs, contrasting the strengths and limitations of UGC with traditional survey data, and advocating triangulation—combining numeric and textual signals, and, where possible, blending UGC with surveys. It also sets out a research agenda that addresses the trade-off between standardization and customization, the management of measurement uncertainty, and the need for validity guidelines for UGC-based CMMs. K