Why Fitness Apps May Be Using the Wrong Exercise Psychology

Most exercise advice starts with information: learn why movement matters, set a goal, track your progress, repeat. That approach is sensible, familiar, and built into nearly every fitness app. But a review of randomized trials in young adults found the most encouraging signals for a different set of tools: material rewards, a valued exercise identity, and habit formation.

That does not mean a step counter belongs in the trash. It means knowledge and targets rarely solve the hard part: doing the workout when work runs late, the weather turns ugly, and motivation has quietly left the group chat.

The evidence points beyond information and goals

A 2021 systematic review and meta-analysis by Whatnall et al. included 66 randomized controlled trials of physical-activity interventions in healthy young adults, generally ages 17 to 35. The authors coded each intervention with the 93-item Behaviour Change Technique Taxonomy v1, then assessed which techniques appeared alongside successful results.

The interventions increased moderate-to-vigorous physical activity, step counts, and total activity in pooled analyses, with effects reported at follow-up periods of three months or less and beyond three months. Those pooled estimates came from much smaller subsets of the full review: 11 trials for moderate-to-vigorous activity, five for steps, and 11 for total activity measured in MET-minutes per week. Only 34 of the 66 trials, or 52%, reported a statistically significant between-group difference.

When the authors looked at individual techniques, three had the highest proportions of studies reporting significant effects: material reward (BCT 10.2), valued self-identity (BCT 13.1), and habit formation (BCT 8.3). Providing information about health consequences (BCT 4.1) and generic goal-setting (BCT 1.1) appeared often but had weaker signals in this narrative comparison.

The distinction makes behavioral sense. Information asks you to make a deliberate choice. Goal-setting asks you to keep revisiting that choice. Habit formation aims to reduce the number of choices required. Identity asks whether exercise fits the person you think you are. Rewards change the immediate payoff of a behavior whose health benefits otherwise arrive on geological time.

This is a directional signal, not a ranking carved in stone

The review does not prove that habit formation, identity, or rewards independently cause larger activity gains than goals or tracking. Its overall statistical test linking the type or number of behavior-change techniques to intervention effectiveness was non-significant. The apparent hierarchy came from a narrative synthesis of the proportion of successful studies using each technique, not from a definitive meta-regression or a head-to-head trial that isolated one technique at a time.

That limitation matters because exercise programs are bundles. A study that includes habit formation may also provide coaching, social support, reminders, structured workouts, and better follow-up. The apparent benefit could reflect the package, the setting, or the people delivering it rather than one coded technique. Different activity measures also complicate the comparison: trials used self-report and accelerometry, different thresholds for moderate-to-vigorous activity, and different step-count protocols.

The short version is more useful than the hype: the review gives people designing an exercise routine a reasonable hypothesis to test, not a validated league table of human motivation.

How to apply the finding without becoming your own behavioral-economics experiment

Use goals and tracking as support tools, not as the entire plan. Then add a mechanism that makes the behavior easier to repeat when enthusiasm dips.

• Build a fixed cue. Attach a small, specific workout to a stable event: walk immediately after lunch, do resistance training after changing into work clothes, or cycle after the last meeting of the day. Keep the cue, place, and first action consistent.
• Make the identity concrete. "I exercise" is stronger operationally than "I should exercise more," but it still needs a behavior attached to it. A person who trains after work on Tuesdays has a script; a person pursuing wellness has a brochure.
• Use a reward with an endpoint. A modest immediate reward can help establish repetition: a favorite podcast reserved for a walk, a social commitment after training, or an agreed consequence for missed sessions. Treat rewards as a starter motor, not proof that the habit will run forever.
• Keep the first version almost embarrassingly easy. The review supports activity interventions, not heroic schedules. A repeatable 15-minute session is more useful than a perfect plan performed twice.

Loss-framed incentives may have appeal because people dislike losing money, but the young-adult review did not establish that loss-framed rewards last longer than gain-framed ones. Do not smuggle a behavioral-economics conclusion into an exercise plan just because it sounds clever.

Rehabilitation studies support the broader adherence argument

Research outside healthy young adults strengthens one narrower point: the support around exercise can change outcomes. In the 2025 OPTIN randomized trial, 126 adults older than 55 with vertebral fragility fractures and back pain received progressive rehabilitation exercises for strength, posture, and balance. Half also received an integrated adherence-support intervention. The supported group had significantly better Timed Up and Go and six-minute walk test outcomes than the standard-rehabilitation group.

That is causal evidence in a clinical setting, but it does not settle which behavior-change techniques work best for a healthy 25-year-old deciding whether to lift weights after class. These participants had fractures, pain, and a strong medical reason to regain function. Their barriers and incentives differ from those of healthy young adults.

A 2024 systematic review of lower-limb fracture rehabilitation reached a compatible, qualified conclusion: behavior-change techniques were associated with better functional recovery and adherence. The review included nine randomized trials, and goal-setting and instruction were among the most common techniques. That is useful context, not a rebuttal of the young-adult review. It shows that a technique can be common and still help within a particular clinical package.

Adherence remains a problem even where exercise has obvious clinical value. A 2025 systematic review and meta-analysis in community-dwelling older adults with sarcopenia, covering 17 randomized trials and 2,975 participants, identified poor adherence as a main factor limiting treatment efficacy. No comparable evidence shows that identity or habit techniques reliably prevent the usual decline in adherence at six to 12 months among healthy young adults.

Activity and sitting are related targets, not interchangeable ones

Adding exercise and reducing sitting often travel together in public-health advice, but they require different plans. An office worker may meet a weekly exercise target and still spend most working hours seated. A systematic review and meta-analysis by Hadgraft et al. examined free-living sedentary-behavior interventions lasting at least seven days. In non-clinical populations, interventions lasting two weeks to less than six months improved cardiometabolic biomarkers. Those biomarkers matter, but they are not the same as patient outcomes or proof of a specific biological mechanism.

Brief "exercise snacks" may offer a practical route for people who cannot protect a single long workout. In a 2026 qualitative pilot study, 20 U.S. office workers who completed an exercise-snack intervention or a self-monitoring condition described barriers and facilitators to participation. The study found that uptake depended on contextual and psychological factors, not merely time. It does not test whether the snacks improve health outcomes, but it explains why a calendar reminder alone may fail when a workplace norm makes standing up feel awkward.

A small 2025 quasi-experimental study in graduate students also addressed the intention-behavior gap. Thirty-two students were randomized to weekly or monthly feedback through an eight-week grid that tracked total exercise volume and aerobic capacity. The intervention improved exercise participation and aerobic capacity. Its size, short duration, and unusual feedback tool limit broad conclusions, but the study addresses the right practical question: how do intentions become calendar entries?

The verdict

Exercise science can prescribe a workout. Behavioral science has to get that workout past Tuesday.

For healthy young adults, the best available synthesis supports a cautious shift in emphasis: pair goals and self-monitoring with a stable cue, an identity-linked routine, and, if useful, a short-term reward. Do not claim that these tools have beaten every alternative in properly powered head-to-head trials. They have not.

The larger unsolved problem is maintenance. The Whatnall review found activity improvements, including some follow-up beyond three months, but it did not establish a behavior-change recipe that reliably preserves gains for six months or longer. Build systems that make exercise easier to repeat, measure whether they actually work for you, and revise the system when they do not. Motivation is a fine guest. It is a terrible landlord.