The potential of User-Generated Content in Marketing Research 

May, 14, 2026

With the rapid and widespread diffusion of Web 2.0 technologies, human users have been, willingly or not, generating and disseminating large amounts of data (402.74 terabytes per day) across digital environments (Duarte, 2024). Consumers actively generate a significant share of this data through user-generated content (UGC), sharing opinions and preferences about brands, products, or services they like or dislike via social media posts, videos, messages, or website visits. In this sense, UGC is a rich data source that provides insight into users’ preferences and consumption patterns, as well as brand and product evaluations. With ever-increasing data availability, novel methods have emerged to analyze large, heterogeneous datasets and provide decision-makers with insights (Kubler et al., 2017). For instance, advances in machine learning (ML) have led to an abundance of models that perform particularly well at predicting and explaining consumer behavior (e.g., Wang et al., 2018; Ringel & Skiera, 2016). Yet, while most ML applications for UGC are effective at predicting consumers’ final decision outcomes (e.g., buy or not buy), their potential to measure latent variables (i.e., unobservable) such as consumer mindset metrics (CMMs) has been neglected (Kubler et al., 2025a). This situation warrants attention for two reasons. First, and most straightforwardly, managers often rely on CMMs as key marketing performance indicators to assess the impact of their marketing efforts. Second, managers still rely on primary data collected through surveys to measure CMMs, despite being subject to several limitations, including high costs, reduced timeliness, susceptibility to sampling error, and response errors (Hulland et al., 2018).

In light of these considerations, we argue that UGC can be a valid alternative to primary data for measuring CMMs.

References