Tabula
ResearchThe appTry it

The research behind Tabula

How to keep a tracker simple enough to stick with

By Alvina · Last updated · 6 min read

Anyone who has downloaded a habit, mood or health app knows how the story usually goes: a burst of enthusiasm, a few weeks of entries, and then a quiet slide into never opening it again. Across health and mental health apps, roughly 70% of people stop within about 100 days (Kidman et al., 2024). It is tempting to read that as a failure of willpower, but the research points somewhere more useful - to design, meaning how often an app asks, how much it asks for, and what happens when a day is missed. This post summarises what studies suggest about each, where the evidence is thin, and how Tabula (a private everything-tracker for iPhone and Mac) responds.

Key takeaways

Why do people stop using tracking apps?

Most people stop early and then keep stopping more slowly. A scoping review of 18 studies (525,824 participants) found a median of 70% of users discontinued within the first 100 days (Kidman et al., 2024).

The same review found a curved pattern: sharp abandonment soon after download, then a slower rate of loss. Reasons fell into six groups - technical and functional issues, privacy concerns, poor user experience, content and features, time and financial costs, and evolving needs and goals. Notably, none of those is "lack of discipline". Independent usage data on 93 popular mental health apps tell a similar story, with a median 30-day retention of just 3.3%, although tracker apps did somewhat better than breathing-exercise apps (Baumel et al., 2019).

Both sources are observational, and a person who stops opening one app may simply have moved on to something that suits them better. Even so, the pattern suggests that the first weeks are where a tracker is won or lost, and that every extra step is a small cost paid early.

How often should a tracking app prompt you?

As rarely as is useful. A systematic review found that more prompts per day, but not more items per prompt, were associated with lower adherence (Machleid et al., 2026).

That review covered 103 studies of ecological momentary assessment and remote monitoring in adults with diagnosed mental disorders, with mean adherence of 74.5% across tracking items. It is not the only finding, though. In an experiment with 163 students, longer questionnaires (60 items rather than 30) increased burden and lowered data quality, whereas asking three, six or nine times a day did not (Eisele et al., 2022). A meta-analysis of 488 ambulatory assessment studies found that a higher total number of measurement occasions and more study days were each related to lower compliance (Ottenstein & Werner, 2022).

Taken together, the safest reading is that total burden matters, whether it arrives as many prompts or as long forms. Which of the two bites harder may depend on the person and the context.

Does asking for less detail make tracking easier?

Usually, yes. Longer forms increased burden in one experiment (Eisele et al., 2022), and low-burden, flexible logging appeared to support awareness at opportune moments (Barker-Canler et al., 2024).

In a field trial of the Chromatize app, low-burden logging of self-defined, symbolically rich data appeared to support self-awareness, reflection and regulation at opportune moments (Barker-Canler et al., 2024). The practical lesson is that a log does not have to be a full report. A single tap, a word or a colour can be enough to mark a moment, with any detail added only if the person wants to.

Why do personal settings matter?

Because adherence seems to rise when an app fits the person. In a systematic review of 99 studies, personalisation or tailoring and individualised push notifications were among four intervention factors linked with better adherence across health areas (Jakob et al., 2022).

The same review also reported that mean adherence across apps was about 56%, and noted that many underlying studies were short pilots. Work with people who have migraine adds a related point: when tracking goals are left unstated, many people struggle to meet them, and interviews suggested that making a goal explicit may help (Schroeder et al., 2018, 2019). In other words, people may track more steadily when they can say what they are tracking for, and choose what to track, what to see and what to be reminded of.

What about low-energy days?

Those are the days a tracker most often loses people, so it helps to have a minimal option. Work on tracking in chronic conditions suggests that people with fluctuating health need a lighter way to log on bad days (Schroeder et al., 2018).

Low-burden, flexible logging can still support awareness at the right moments (Barker-Canler et al., 2024). A tracker that demands the same effort on a hard day as on a good one quietly teaches people that tracking is only for good days, which is exactly when data is least informative.

What happens when you miss a day?

Ideally, nothing punishing. Lapsing and resuming appears to be a normal part of personal tracking (Epstein et al., 2015), and one study of habit formation found that missing a single opportunity did not materially affect the process (Lally et al., 2010).

In that study, 96 volunteers repeated an everyday behaviour for 12 weeks, and the time to reach a plateau of automaticity ranged from 18 to 254 days. Those were eating, drinking and activity habits, not logging itself, so it is a loose analogy rather than direct proof. A separate line of experiments found that treating oneself with compassion after a failure increased motivation to improve, somewhat paradoxically (Breines & Chen, 2012). Together these suggest that a guilt-heavy response to a gap, such as a red empty day or a broken streak, is unlikely to be the best way to bring someone back.

What is missing from the evidence?

A lot, especially for the people a tracker is for. Most studies measure adherence in general, often in trials with reminders from researchers, so they say less about someone choosing to log for themselves for months.

Several results also pull in different directions. The frequency and length findings above do not fully agree, the student sample may not generalise, and the habit study was not about logging. Head-to-head comparisons of streaks and forgiving designs appear rare, and little of this work focuses on neurodivergent users or people with fluctuating energy. The best setup likely differs from person to person, which is why Tabula makes nearly everything optional rather than claiming a single correct rhythm.

What does Tabula do about it?

Tabula tries to keep each log small and each gap harmless. It has no streaks, minimal one-tap logging, reminders you choose and can snooze, and a gentle welcome back with easy backfill instead of red empty days.

There is a minimal "low-capacity day" log for hard days, and Oratio lets you say or type a messy day and check what it understood before anything is saved. Check-ins use emotion words rather than demanding a number, and your data and the on-device AI stay on your device. Tabula is a personal journal, not a medical tool or a treatment, and it makes no claim that tracking improves any condition. Whatever benefit tracking has depends on people actually using it, which is the point of keeping it light.

References

Barker-Canler, M., Gooch, D., Van Der Linden, J., & Petre, M. (2024). Flexible minimalist self-tracking to support individual reflection. ACM Transactions on Computer-Human Interaction, 31(3), 1–35. https://doi.org/10.1145/3660339

Baumel, A., Muench, F., Edan, S., & Kane, J. M. (2019). Objective user engagement with mental health apps: Systematic search and panel-based usage analysis. Journal of Medical Internet Research, 21(9), Article e14567. https://doi.org/10.2196/14567

Breines, J. G., & Chen, S. (2012). Self-compassion increases self-improvement motivation. Personality and Social Psychology Bulletin, 38(9), 1133–1143. https://doi.org/10.1177/0146167212445599

Eisele, G., Vachon, H., Lafit, G., Kuppens, P., Houben, M., Myin-Germeys, I., & Viechtbauer, W. (2022). The effects of sampling frequency and questionnaire length on perceived burden, compliance, and careless responding in experience sampling data in a student population. Assessment, 29(2), 136–151. https://doi.org/10.1177/1073191120957102

Epstein, D. A., Ping, A., Fogarty, J., & Munson, S. A. (2015). A lived informatics model of personal informatics. In Proceedings of the 2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing (pp. 731–742). ACM. https://doi.org/10.1145/2750858.2804250

Jakob, R., Harperink, S., Rudolf, A. M., Fleisch, E., Haug, S., Mair, J. L., Salamanca-Sanabria, A., & Kowatsch, T. (2022). Factors influencing adherence to mHealth apps for prevention or management of noncommunicable diseases: Systematic review. Journal of Medical Internet Research, 24(5), Article e35371. https://doi.org/10.2196/35371

Kidman, P. G., Curtis, R. G., Watson, A., & Maher, C. A. (2024). When and why adults abandon lifestyle behavior and mental health mobile apps: Scoping review. Journal of Medical Internet Research, 26, Article e56897. https://doi.org/10.2196/56897

Lally, P., van Jaarsveld, C. H. M., Potts, H. W. W., & Wardle, J. (2010). How are habits formed: Modelling habit formation in the real world. European Journal of Social Psychology, 40(6), 998–1009. https://doi.org/10.1002/ejsp.674

Machleid, F., Michnevich, T., Huang, L., Schröder-Frerkes, L., Wiegmann, C., Muffel, T., & Kaminski, J. (2026). Remote measurement-based care interventions for mental health: Systematic review and meta-analysis. JMIR Mental Health, 13, Article e63088. https://doi.org/10.2196/63088

Ottenstein, C., & Werner, L. (2022). Compliance in ambulatory assessment studies: Investigating study and sample characteristics as predictors. Assessment, 29(8), 1765–1776. https://doi.org/10.1177/10731911211032718

Schroeder, J., Chung, C.-F., Epstein, D. A., Karkar, R., Parsons, A., Murinova, N., Fogarty, J., & Munson, S. A. (2018). Examining self-tracking by people with migraine: Goals, needs, and opportunities in a chronic health condition. In Proceedings of the 2018 Designing Interactive Systems Conference (pp. 135–148). ACM. https://doi.org/10.1145/3196709.3196738

Schroeder, J., Karkar, R., Murinova, N., Fogarty, J., & Munson, S. A. (2019). Examining opportunities for goal-directed self-tracking to support chronic condition management. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 3(4), 1–26. https://doi.org/10.1145/3369809