Forgetting and Not Starting Are Different Problems (2026)

Last updated: 2026-09-07

Quick answer: In our review corpus of 32,021 habit app reviews, the theme we call forgetting shows a 0.55 star gap between Routinery and the other 12 apps (4.30 against 3.74, 284 reviews). The theme we call failing to start shows a smaller 0.26 star gap (4.13 against 3.87, 145 reviews). Habit research treats these as separate mechanisms too. A missed cue is a memory failure that an external reminder can fix. A routine that never launches is a planning gap that a reminder does not reach.

Two of the nine complaint and praise themes we track in these reviews describe someone who did not do the thing that day. One reviewer says the app never reminded them, or that they simply forgot it was there. Another says they downloaded the app, opened it once, and never went back. Our review data counts these separately, and the two gaps are not close to the same size.

What we counted: two themes inside 32,021 reviews

We pulled two of nine complaint and praise themes out of 32,021 public Google Play reviews across 13 habit, routine and task apps, collected between 2018-12-22 and 2026-08-19, up to 3,000 per app. Routinery is one of the 13, measured with the same phrase sets and the same per-app cap as the rest.

Forgetting is matched against phrases like forget, forgot and slip my mind. Failing to start is matched against a separate phrase set describing an app that was installed but never got past the first session. A review counts once per theme, and no individual review is quoted anywhere in this article.

Two limits apply to both counts. Phrase matching only catches wording we anticipated, so a reviewer describing either experience in unusual terms gets missed, which makes every number here a floor. And this is Android data only. The Apple review feed has returned empty results for these apps since August 2026, and we are flagging that gap here instead of leaving it out.

The gap: 0.55 stars against 0.26 stars

Forgetting separates Routinery from the field by more than double what failing to start does.

Both rows move the same direction. Reviewers who bring up either theme rate us above the rest of the field on it. The overall average across all 32,021 reviews is 3.96 stars, and a forgetting complaint about any app in the set tends to sit below that line while a failing-to-start complaint sits closer to it. Our own samples on both rows are small, 54 and 31 reviews, so we treat 4.30 and 4.13 as directional estimates.

A single average could tell you that reviewers on both themes prefer us. It would not tell you that forgetting and failing to start describe two different moments in someone's week.

Why the theory backs the split: difficulty and context are not the same variable

Habit-formation research already treats difficulty and context as separate ingredients, and that lines up with what the review counts show. A 2026 preregistered reanalysis followed 254 adults over 6 months and tested 10 candidate factors against habit strength. 6 of the 10 factors were significantly linked to it, five in the positive direction and one negative: perceived difficulty. Context stability, a measure of how steady the setting and timing of the behavior stayed, sat among the positive six.

Difficulty and context landed in the same statistical model as two separate variables. One is about how hard the action feels to begin. The other is about whether the cue that used to trigger it is still where it was. The study is an observational reanalysis of data gathered during a period with unusual disruption to routines, so it shows association. A reanalysis like this cannot prove that either factor causes habit decay on its own.

Fixing forgetting: an external signal beats memory

Forgetting responds to a signal that arrives from outside your own memory. A systematic review and meta-analysis of 14 randomized trials on medication reminders found an average adherence improvement of 0.57 points on the MMAS-8 scale where a reminder feature was present, and 12 of the 14 apps studied offered one.

That trial population takes daily medication for chronic conditions, a different domain from habit tracking, so we treat the direction as the takeaway and set the exact figure aside. The direction still matches what a forgetting complaint describes: a moment that needed a nudge from outside and didn't get one. No amount of intending to remember replaces that nudge once the cue has already passed.

Fixing failing to start: a plan made before the moment arrives

Failing to start is the gap between deciding to do something and doing it. A meta-analysis of 94 studies on implementation intentions, plans that tie a specific cue to a specific action ahead of time, found an average effect size of 0.65, a medium to large effect by the standards of behavioral research.

The mechanism differs from a reminder in an important way. A reminder supplies the cue after the fact, once the window to act is already open. An implementation intention builds the cue into the decision before the moment shows up, so there's nothing left to remember when it does. That's also why stacking more reminders onto a failing-to-start problem tends to disappoint. The reviewer isn't missing a nudge. They never built the if-then link in the first place.

Which one you have: three checks before you blame the app

The two problems look alike from the outside, so check which one applies before switching apps or turning on more alerts.

Look at whether a notification fired at all. If your phone shows a missed reminder and you scrolled past it, that's forgetting. If no notification exists for that habit yet, or you turned reminders off weeks ago, the problem sits upstream of any reminder feature.

Pull up a week of your own activity log and check where the gaps sit. Forgetting tends to scatter across days you were otherwise on track. Failing to start tends to cluster around habits you added and never opened once, regardless of the day.

Ask what happens right before the habit is supposed to happen. If you can name the moment, the location and the action in one sentence ahead of time, you likely have a forgetting problem a reminder can patch. If you can't, the plan itself is missing, and a louder reminder won't build it for you.

Routinery: measured on the same terms, with a caveat

We make Routinery, so the 4.30 and 4.13 figures above come with a caveat we want to state plainly. Both rest on small samples, 54 and 31 reviews, against a field comparison of 230 and 114. We treat both as directional until the sample grows.

What we can say with more confidence is that the two theories point to different design choices, and we've built toward both instead of picking one. A scheduled reminder tied to a specific time addresses the forgetting side. A routine that asks you to set the trigger moment before you start it, not after, addresses the planning side. We've also written about what to track first when you start a habit app, since the earliest days carry the highest risk for both themes before a routine has had time to stick.

The split matters outside our own numbers too. Our review of setup burden in ADHD habit app reviews found a different theme, setup difficulty, driving most of the complaints for that group. The lesson carries over: check which theme actually applies to your case before assuming reminders are the fix, because for some readers they aren't.

Frequently asked questions: forgetting and not starting in habit apps

Are forgetting and not starting the same habit problem?

Not in our data or in the research we found. Forgetting shows a 0.55 star gap between Routinery and the rest of the field across 284 reviews, while failing to start shows a 0.26 star gap across 145 reviews, roughly half the size. Habit-strength research backs the split too. A study of habit strength determinants found perceived difficulty and context stability as two of six factors that moved habit strength independently. Forgetting is a missed cue. Failing to start is a plan that was never built. They call for different fixes, and treating them as one problem tends to leave one of the two unaddressed.

What counts as forgetting versus failing to start in this review data?

Forgetting reviews use language like forget, forgot or slip my mind, describing a routine the reviewer had been doing before a specific instance got missed. Failing-to-start reviews describe an app that was downloaded, opened once or twice, and never became a habit at all. The two phrase sets don't overlap, and a review is counted in only one theme. Both are drawn from the same corpus of 32,021 public Google Play reviews across 13 apps, collected through 2026-08-19.

Does a reminder fix a failing-to-start problem too?

Not on its own, based on the research we cite above. A meta-analysis of 94 studies found implementation intentions, plans that link a cue to an action before the moment arrives, produce a 0.65 effect size on follow-through. A reminder arrives after the decision point has opened, which helps when the plan already exists and the issue is memory. When the plan itself is missing, which is what a failing-to-start complaint usually describes, an extra alert has nothing to attach to.

How big is the ratings gap for each theme, and how reliable is it?

Forgetting shows Routinery at 4.30 against 3.74 for the rest of the field, a 0.55 gap across 284 total reviews. Failing to start shows 4.13 against 3.87, a 0.26 gap across 145 total reviews. Our own samples behind those two numbers, 54 and 31 reviews, are small enough that we treat 4.30 and 4.13 as directional estimates. The comparison figures for the other 12 apps, built from 230 and 114 reviews, rest on a larger base.

Is this review data representative of everyone who uses these apps?

Treat it as a signal about what public reviewers write. It is not a full picture of every user. These are Android reviews only, since the Apple feed for these apps has returned empty results since August 2026. People who leave public reviews also skew toward strong opinions in both directions, not a random sample of everyone with the app installed. Phrase matching misses wording we didn't anticipate, so every count in this article sits on the low side of what actually happened.

Method: how the counting was done

Method. 32,021 public Google Play reviews across 13 habit, routine and task apps, up to 3,000 most recent reviews per app, collected 2026-09-01. Routinery is one of the 13 and is measured with the same phrase sets and sampling rule as the rest. Reviews were matched against two fixed phrase sets: forget, forgot and slip my mind for the forgetting theme, and a separate set for reviews describing an app abandoned before it became a routine. A review counts once per theme. No individual review is quoted in this article.

Limits. Google Play only. Public reviewers are not a random sample of all users. Phrase matching undercounts reviews that describe either theme in wording we did not anticipate. Both Routinery samples in this article, 54 and 31 reviews, are small enough to read as directional.

References: peer-reviewed sources cited above

• Bürgler S, Milyavskaya M, McMillan G, Burns RJ, Lally P, Gardner B. What Makes a Habit? Investigating Potential Determinants of Habit Formation. Personality and Social Psychology Bulletin, 29 April 2026. n=254, preregistered reanalysis.
• Lanke, Trimm, Habib, Tamblyn. Evaluating the Effectiveness of Mobile Apps on Medication Adherence for Chronic Conditions: Systematic Review and Meta-Analysis. Journal of Medical Internet Research, 31 July 2025. 14 randomized trials, n=1,565.
• Gollwitzer PM, Sheeran P. Implementation Intentions and Goal Achievement: A Meta-Analysis of Effects and Processes. Advances in Experimental Social Psychology, vol 38, 2006. 94 studies.

About the author. Written by the Routinery team. Routinery pairs a scheduled reminder with a routine that asks you to set the trigger moment in advance, which makes us an interested party in the sections above about reminders and planning.

Reviewed by the Routinery product team. No individual review is quoted in this article, no reviewer is identified, and raw review text is not published. Every figure comes from an automated count.