Do habit trackers actually work?

Updated 19 August 2026

• Self-monitoring works, and it does not require an app. Paper does it too.
• Naming when and where you will act has stronger evidence than any tracking feature.
• The reward layer is the least-evidenced part and the most heavily marketed.
• A tracker cannot tell you why a habit stopped unless it asks, and most do not.
• An app that makes a missed day expensive is working against the thing you installed it for.
• If your routine already works, a tracker adds very little. It earns its place when habits keep failing.

What has evidence behind it

Self-monitoring. Keeping track of a behaviour is one of the more consistently effective techniques in behaviour-change research, and it turns up as an active ingredient across a wide range of interventions. This is the honest core of the whole category. Note that it says nothing about apps: a paper chart does it.

Naming when and where. Forming a specific plan of the form "when situation X arises, I will do Y" had a medium-to-large effect across a meta-analysis of 94 studies. This is stronger evidence than anything specific to tracking, and it is a setup question rather than a feature.

Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69–119.

Repetition in a stable context. The mechanism itself. Automaticity rises with repetitions in a consistent setting, on a curve with a median of 66 days and enormous individual spread.

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. doi:10.1002/ejsp.674

What does not, or is weaker than the marketing

Consecutive-day counters. Effective while intact and actively harmful when they break, because they convert a day with no measurable cost into a visible loss. In the field data, one missed opportunity did not measurably affect habit formation — so the loss is manufactured by the product.

Points, levels and badges. They work well for people who are motivated by games and add friction for everyone else. There is also a known risk that an external reward becomes the reason you do something, which is fine until the reward stops being interesting.

Charts and completion percentages. Mostly self-monitoring in a costume, with the same loss problem as a streak. A falling percentage arrives precisely when support would be more useful than a number.

Reminders. Genuinely useful early as scaffolding, and a common way to end up with a habit of dismissing notifications rather than one of doing the thing.

So do they work?

A tracker helps if it does the evidenced things and avoids the rest:

• Gets you to name a specific behaviour, a specific moment and a specific place.
• Makes recording cheap enough to happen daily.
• Collects why a day did not happen, in a form it can count.
• Does not make a missed day expensive.
• Proposes a change to the plan when a pattern appears.

Most of that list is setup and questions rather than features, which is why a notebook does surprisingly well and why an expensive app is not obviously better than a cheap one.

When a tracker is the wrong tool

• Your routine already works. Self-monitoring has least to offer where behaviour is already stable. You want reminders, and those are free.
• You have never kept a habit for a month. The problem is more likely the size of the habit or the absence of a bad-day version, and both are free to fix.
• You want the app to supply motivation. It will for about eleven days.

The honest summary

Tracking works, modestly, and mostly through the plainest thing it does: making you notice. The features that get marketed hardest are the least supported, and one of them — the counter — is actively counterproductive at the moment that decides whether a habit survives.

Which means the useful question is not whether habit trackers work but whether a particular one does the evidenced things without the harmful one. That is answerable in about two minutes on any app's screenshots.