Conference Agenda

Digital platform ecosystems rely heavily on engagement-based monetisation, making user attention a core resource that platform owners steer through design and algorithmic curation. In this context, so-called addictive design features, such as algorithmic personalisation and infinite scrolling/autoplay, can exploit psychological biases, reduce user autonomy, and contribute to problematic, potentially addictive use. This raises critical attention governance questions that have also become salient in regulations such as the Digital Services Act and the announced Digital Fairness Act. Against this regulatory backdrop, this study examines such features as attention-steering mechanisms and collects new evidence on users' perceptions of the impacts of addictive design features under varying circumstances on two dimensions: service-specific time effects and cross-service purchasing effects, and implications for policy design.
We address three exploratory research questions: (1) How do users perceive addictive design features, and how does this perception relate to addictive behaviour in terms of time spent on the platform and propensity to purchase? (2) What mechanisms do users utilise to regulate their behaviour? (3) How do perceived effects and self-regulation differ by platform type?

The analysis draws on an online consumer survey conducted in Germany in November/December 2025 (N = 3,252), with quota sampling for age, gender, and region for respondents aged 16+. Respondents were allocated to a between-subject design conditional on usage, covering social networks (N = 1,049), video-sharing platforms (N = 1,042), and online shopping services (N = 1,043). Six feature categories, adapted to each service category, are examined: continuous content supply (infinite scrolling/autoplay), personalised content, gamification, visibility of likes/ratings, notifications, and ephemeral content. For features to which respondents reported exposure, the perceived effects were elicited using 5-point scales for (i) usage time of the respective service category and (ii) general online shopping behaviour. To capture service-specific and shopping addiction tendencies, the study employs adapted versions of the Bergen Social Media Addiction Scale and the Bergen Shopping Addiction Scale. Furthermore, the survey elicited information on past use of self-regulation strategies and willingness to use them in the future.

The results of the study indicate significant heterogeneity in user perceptions. However, several consistent patterns emerge, from which the following key findings are derived.
The initial key finding establishes that, across all feature categories, approximately half of respondents report no perceived effect on time spent, and this proportion is somewhat larger for effects on online purchasing. Concurrently, meaningful fractions report strong effects in both directions, with the fraction perceiving time-prolonging effects being generally larger than the opposite fraction, and the reverse being true for online shopping. On average, infinite scroll/autoplay and personalised content are perceived to exert the strongest significant time-prolonging effects, whereas gamification and ephemeral content are closer to a neutral average effect. Concerning online shopping, most features are perceived as having a slightly purchase-reducing effect, with personalised content representing the main exception.
A second key finding is that stronger perceived feature effects are associated with significantly higher shares of respondents above addiction-risk thresholds. While the data does not establish causality, the observed pattern is consistent with the hypothesis that addictive design features contribute to the development of addiction.
Thirdly, the capacity for self-regulation appears limited. The majority of respondents reported neither past efforts to regulate their use, nor a perceived need to do so in the future. Even among higher-risk users, only a minority reported having adopted behavioural controls. Where self-regulation does occur, it most often targets salient cues mechanisms as notifications, alongside time-tracking tools and, in some cases, temporary or permanent deactivation. Fourthly, the aggregated effects are predominantly reflected when differentiating by service group. However, the effects on social networks are generally perceived to be slightly more time-prolonging. Furthermore, certain features do not consistently achieve significance across services, and, for ephemeral content, even a change in effect direction occurs. Regarding online shopping behaviour, the impact on online shopping services is generally the least purchase-reducing.

Overall, this study demonstrates that, despite the evident heterogeneity, a considerable proportion of respondents perceive addictive design elements to influence their usage time. Considering existing literature on this subject, the actual influence may be even larger due to perception bias. Furthermore, tendencies towards digital addiction are evident in the sample, with more pronounced tendencies among respondents who perceived stronger feature effects. These findings raise further concerns about whether digital platforms exploit their attention-steering capabilities to their own advantage, but at the expense of consumer well-being. At the same time, consumers are not fully aware of the risks and harms involved, and only partially utilise the limited self-regulation mechanisms available to them.