Your Fitness App Is Missing the Point: New Research Reveals Why Personalization Fails

Millions of adults strap on smartwatches and download fitness apps every year, convinced that a step counter and a heart rate monitor are all they need to get moving. Yet most of these devices end up abandoned in drawers within weeks, and the global burden of physical inactivity keeps growing. A new study from Mälardalen University in Sweden, published in Frontiers in Digital Health, argues that the problem is not the technology itself but a fundamental misunderstanding of what personalization actually means. The researchers, Agnieszka Jaff, Mia Folke and Annica Kristoffersson, show that effective personalization must reach far deeper than demographics and step counts, into what users know, how they feel, and how they actually live their lives.

The World Health Organization lists insufficient physical activity as one of the leading risk factors for global mortality, contributing to cardiovascular disease, diabetes, certain cancers and reduced quality of life. Mobile technologies have emerged as accessible and cost-effective tools to encourage healthier behaviors, offering scalability, real-time feedback and rapidly evolving capabilities. Personalization is widely recognized as a key factor for improving user engagement and adherence, and systematic reviews of mobile physical activity apps show that effective interventions typically include individualized goals, tailored feedback and context-sensitive reminders. But current approaches, the researchers argue, often rely on surface-level data and overlook the cognitive and motivational aspects that truly drive behavior change.

To bridge this gap, the team turned to an unexpected theoretical ally: the Knowledge, Attitude and Practice model, or KAP, a framework long used in healthcare to understand and influence health-related behaviors. The model examines what individuals know about a subject, their feelings toward it, and their actions relating to it. The researchers integrated this model with human-centered design, a methodology that begins with building design empathy—viewing the world from another person’s perspective by setting aside one’s own assumptions. Developers, they note, often rush into solving the how of design without fully understanding the why, clarifying human needs and motivations before focusing on the final design outcome.

The study involved semi-structured interviews with 24 adults aged 34 to 66 years, recruited through academic networks in Sweden. Thirteen interviews were performed directly after an assessment of physical abilities related to fall risk, and 11 were performed after participants had followed a 4-week individual exercise plan. The interviews, conducted in Swedish or English and lasting between 11 and 21 minutes, explored participants’ physical activity habits and their use of exercise- and health-related technology. Recordings were transcribed and analyzed using deductive thematic analysis guided by the three KAP domains, with themes iteratively refined and reviewed throughout the process. The study was approved by the Swedish Ethical Review Authority.

The knowledge domain revealed striking variation. Most participants associated physical activity with general well-being and improved mood, often based on personal experience rather than scientific understanding. Only a few demonstrated broader, evidence-informed awareness of preventive and long-term health effects. Crucially, general awareness of benefits was not always accompanied by practical knowledge about planning exercise or interpreting activity data. One participant admitted, “I haven’t really figured it out, I must say. But I do understand that there are big advantages with measuring heart rate, distance, and things like that. But I’ve actually never done it.” Others collected data passively, out of curiosity, without using it in a structured feedback loop. Personal experiences—menopause, weight gain, physical injury—often fostered more nuanced understanding than formal education, suggesting that lived experience is a powerful form of knowledge that apps rarely tap into.

The attitude domain uncovered equally complex emotional terrain. Health maintenance emerged as the dominant rationale, with participants framing physical activity as essential for sustaining function and independence rather than pursuing performance or appearance. “It’s more about not getting worse rather than getting better,” one participant explained. Intrinsic motivators—enjoyment, emotional regulation, stress relief—acted as strong forces, with some describing physical activity as a form of therapy. But negative emotions ran deep as well: several participants reported guilt or stress related to their use of tracking technology. “It only made me feel bad … I felt guilty,” one participant said of a step counter. Another noted, “The way my mind works makes me more stressed by digital tools. So, it’s calmer for me to work out without them.” These ambivalent reactions underscore that personalization requires sensitivity to both motivating and demotivating features, offering users control over how feedback and reminders are delivered.

The practice domain showed that physical activity is deeply embedded in daily routines, contexts and constraints. Participants combined structured exercise—strength training, yoga, running, swimming—with unstructured activities such as caregiving, commuting, gardening and occupational demands. Walking the dog, biking to work on an e-bike, and taking 20,000 steps during a work shift all counted as meaningful movement. Social preferences were equally diverse: some valued companionship and accountability, while others prioritized autonomy and solitude. Use of technology ranged from reliance on detailed data analytics, with one participant exporting statistics into Excel to compare correlations across graphs, to deliberate avoidance. Structural barriers—limited time, physical limitations, weather—were often more decisive than motivational ones.

Synthesizing these insights, the researchers produced a framework of 20 dimensions distributed across the knowledge, attitude and practice domains, each capturing a specific aspect of how individuals relate to physical activity and technology. The dimensions include awareness of physical activity’s effects, familiarity with mobile health technology, experiential knowledge, identity and self-perception, intrinsic motivation, progress orientation, social influence, internal barriers, structured physical activity engagement, everyday activity integration, technology use, environments, temporal strategies and barriers affecting practice. Each dimension carries distinct intra-dimensional variations—for example, a user’s familiarity with technology may range from unfamiliar to applied understanding, where data insightfully guides behavior. Recognizing these variations allows developers to tailor content, feedback and behavior change strategies to users’ distinct knowledge levels, motivational drivers and practices.

The findings align closely with principles from the Persuasive Systems Design framework, particularly tailoring, social facilitation and normative influence, but the researchers argue that established design principles require expansion to capture the diversity they observed. Prior studies such as the JitaBug app for older adults and the Active2Gether intervention incorporated attitudinal factors and contextual data, yet none has assessed users’ physical activity knowledge, attitudes and practices combined as personalization inputs. Sullivan and Lachman’s review noted that fitness technology rarely targets cognitive dimensions, focusing instead on surface behaviors, and that many users abandon devices within weeks while struggling to interpret data meaningfully. The new study directly addresses this critique by arguing that both cognitive and behavioral data about users should inform personalization models.

The implications are concrete. Fitness apps could provide accessible education for those with limited scientific knowledge, motivational reinforcement for those struggling with self-discipline, or functional framing for those prioritizing health maintenance with age. For users with limited exercise-planning knowledge, apps could suggest short activity sessions at suitable times and make them easy to reschedule. Feedback complexity and frequency should be adaptable, enabling users to reduce reminders that trigger guilt. And everyday movement—caregiving, occupational demands, active transport—should be recognized and supported rather than narrowly promoting formal workouts. The researchers caution that the framework should not require all dimensions to be assessed simultaneously; future work could explore progressive profiling or ecological momentary assessments to gather relevant information over time. Intervention studies, dynamic adaptation to changing motivation and life circumstances, and ethical considerations around data collection and user control remain essential next steps. But the core message is clear: one-size-fits-all approaches are inadequate, and technologies that resonate with users’ identities, motivations and daily contexts are far more likely to foster meaningful and lasting behavior change.

Subject of Research: Health

Article Title: Informing personalization of technologies for promoting physical activity: bridging the knowledge, attitude and practice model and human-centered design

Article References: Jaff, A., Folke, M., & Kristoffersson, A. (2026). Informing personalization of technologies for promoting physical activity: bridging the knowledge, attitude and practice model and human-centered design. Frontiers in Digital Health, 8, Article 1959127. https://doi.org/10.3389/fdgth.2026.1959127

Image Credits: AI Generated

DOI: 10.3389/fdgth.2026.1959127

Keywords: personalization, physical activity, mobile health, KAP model, human-centered design, behavior change, wearables, digital health, Informing, technologies, promoting, physical