# Fell Off a Habit? Diagnose the Lapse, Then Restart Small

> Treat a lapse as data: a five-cause diagnosis table, a restart date no more than three days out, a smaller version, and a written plan for the next miss.

- Canonical page: https://learnhubplay.nl/blog/restart-a-habit-after-a-lapse
- Published: 2026-10-23
- By: LearnHubPlay, an independent software studio in Amsterdam
- Topic: Habits & Routines
- Tags: Habit formation, Habit tracking, Behaviour change, Troubleshooting

You were walking after dinner, or writing before work, or reviewing flashcards, and then a trip, a deadline or a cold took three days. Then ten. Now it is week six, and the plan is either "when things calm down" or "back to the full routine on Monday." Both skip the same question: why it stopped.

This guide treats the gap as data: separate a skipped day from a real lapse, check five common causes, pick a restart date, re-enter at a size that survives your worst day, and write a plan for the next miss.

**The short answer.** One skipped day barely registered in Lally and colleagues' 2010 habit study; a week without the behaviour is a different case. Relapse-prevention research proposes that your reaction to a slip helps decide whether it becomes a relapse, though the evidence is mixed. So name the cause, restart at the tiny version you would do on your worst day, begin at tomorrow's cue (or a week or month start no more than three days away), and write down what you will do when you miss again.

## A miss, a lapse, and a relapse are three different events

Relapse-prevention research began in addiction treatment, and its vocabulary fits habits better than "failure" does. Larimer, Palmer and Marlatt's 1999 review of Marlatt and Gordon's model (*Relapse Prevention*, 1985) defines a lapse as "the first violation of the abstinence goal" and a relapse as a return to uncontrolled use or abandonment of the goal, and says "a critical difference exists" between them.

In the model, a lapse becomes a relapse partly through the reaction to it, the abstinence violation effect, which "may influence whether a lapse leads to relapse." Blaming personal failure brings guilt, which can feed further lapses; blaming stable causes beyond your control ("I have no willpower") makes giving up likelier; seeing a specific coping failure makes learning from it likelier. Dieting researchers describe a cousin, the "what-the-hell effect": the diet is already broken, so why not carry on.

The evidence is mixed. The review cites Miller and colleagues (1996): participants most committed to absolute abstinence, and so likeliest to feel guilty about a lapse, were likeliest to relapse. Among 203 smokers with 1,001 lapses, though, self-blame, self-efficacy and guilt after the first lapse did not predict relapse, and after later lapses more self-blame went with a slower slide to the next one while a drop in confidence went with a faster one ([Kirchner, Shiffman and Wileyto, 2012](https://pmc.ncbi.nlm.nih.gov/articles/PMC3296289)). The what-the-hell effect was not consistently replicated in follow-up studies (Adriaanse and Ten Broeke, 2022). Treat both as well-described ideas with mixed support, not laws.

> **Note:** Both ideas come from quitting-type goals, so applying them to a new habit is an analogy. If your lapse was staying off alcohol, nicotine or other substances, this guide is not treatment; contact a clinician or a quit line.

### Where a miss ends and a lapse begins

In Lally and colleagues' 2010 study of 96 volunteers building a daily habit, 140 missed opportunities (across 55 people, each right after three straight days of doing it) lowered automaticity by an average of 0.29 points on a 42-point scale. The authors concluded that one miss "did not materially affect the habit formation process." Only 67 of the 140 had a next-day score; [how long it takes to form a habit](https://learnhubplay.nl/blog/how-long-to-form-a-habit) has the numbers.

In Armitage's 2005 exercise study a lapse meant a week without attending, and lapses predicted worse later performance. Lally's team reads the two together: one miss does not preclude a habit, but a week of misses hinders it. Every miss Lally counted was isolated, and "a week" is the smallest lapse Armitage could see, not a discovered threshold. No study we found tests the stretch between, so the table is our working rule, with James Clear's "never miss twice" (see [the habit-tracking guide](https://learnhubplay.nl/blog/habit-tracking-that-sticks)) as its second row.

| Days since you last did it | What it is | What to do |
|---|---|---|
| 1 | A miss | Tiny version at the next cue; no make-up session |
| 2–6 | An early lapse | On the second miss, write one line naming the cause |
| 7–30 | A lapse | Run the re-entry protocol below |
| 31–90 | A long lapse | Run the protocol, then check the goal still matters |
| Over 90 | A new habit | Fresh cue and tiny version; count 14 days from day one |

## Self-compassion after a lapse: what four experiments show, and where they stop

Some people answer a lapse with self-criticism, on the theory that guilt keeps you honest. In four experiments, Breines and Chen (2012) compared self-compassion after a mistake with control conditions. The self-compassion group reported more motivation to make amends for a moral transgression and studied longer for a difficult test after a failure, and the authors concluded that, "somewhat paradoxically," an accepting approach to failure may make people more motivated to improve.

There are limits: these were short laboratory tasks, not habit lapses, and they measured motivation within a session, not weeks of behaviour. Our inference, which none of these studies tested, is to name a specific cause, which keeps accountability (self-blame did not speed the smokers' slide), without turning it into a verdict on yourself, which protects confidence (a drop in confidence did). "My mornings changed and a 45-minute run no longer fit" points at something you can change; "I'm hopeless" does not.

## Diagnose the lapse: five causes and what each asks you to change

Spend ten minutes with a pen before choosing a plan; a diagnosis in your head tends to end at "I'm lazy."

```text
LAPSE LOG
Last day I did it: ______      Days since: ______
What changed in the 7 days before that? ______
At the usual cue time on the last three skipped days, I was: ______
What I told myself about the gap: ______
  Rewritten as a specific cause I can change: ______
Version I last did: ____ min.   Version I could do on my worst day: ____ min.
Would I make the same choice again?  yes / no
```

These five causes are our working categories, not a validated typology.

| Likely cause | How to recognise it | What to change |
|---|---|---|
| **Context changed** (schedule, move, trip, people) | The last good day sits next to a change of place, time or company | Drop the old cue; restate the plan around a daily event in your current routine |
| **Habit too big** | You did it only on days with spare time and energy | Cut to a version of two minutes or less that counts as done |
| **No clear cue** | You cannot finish "after I ___ at ___" in one line | Pick one daily event and write an [if-then plan](https://learnhubplay.nl/blog/implementation-intentions-if-then-plans) |
| **No reward you feel** | You did it because you should and never looked forward to it | Change the form: somewhere you like, a podcast or coffee only during it, or with another person |
| **Competing priority** | You had the time, chose something else, and would again | A decision, not a habit problem: pause it with an end date written down now, or retire it |

Not on the list: illness, injury, grief or low mood. Restart at the tiny version once you have the capacity, and ask a clinician first if the cause is lasting or medical.

One row has research behind it. Wood, Tam and Witt (2005) measured the exercise, newspaper-reading and TV habits of 115 students a month before and after they transferred universities. Strong exercise habits survived the move when students kept exercising in the same location (home, gym) and faded when it changed, so in our reading a changed context is where an explicit new plan matters most. It is one study of self-reported behaviour. The size row is a rule of thumb: Lally's exercisers took longer to plateau than drinkers and eaters, but not significantly (p = .328).

## Pick a restart date: tomorrow, or a landmark no more than three days away

Dai, Milkman and Riis (2014) analysed Google searches for "diet," visits to a university gym by 11,912 undergraduates over 442 days, and goal contracts on stickK. In the gym data, relative to baseline, the probability of a visit rose by:

| Landmark | Change in the probability of a gym visit |
|---|---|
| Start of a semester | +47.1% |
| Start of a week | +33.4% |
| After school breaks | +24.3% |
| Start of a month | +14.4% |
| Start of a year | +11.6% |
| A birthday other than the 21st | +7.5% |

The authors suggest landmarks "relegate past imperfections to a previous period," which is what a lapse needs.

Three caveats. These are records of attendance, searches and sign-ups around a date, not of whether restarters kept going. Their plots suggest the lift fades, fastest after a holiday. And a 2026 meta-analysis of 31 studies (N = 13,765) of landmarks' affective, cognitive and self-regulatory outcomes in consumption and decision-making contexts found a small average effect (g = 0.33, 95% CI 0.05 to 0.61) with I² of 90.8%, which its authors call a context-dependent average tendency, not a stable effect.

Our rule: plan day one for tomorrow's cue. If the coming Monday or the 1st of the month is three days away or less, make that day one instead, and do the tiny version in between if you can; if it is further off, start tomorrow and treat that Monday as your first weekly check. No study we found compares waiting with starting at once, and Dai and colleagues themselves flag that people may delay plans until a landmark, leaving that to future research, so the three-day cap is our judgement.

## The re-entry protocol: six steps and a worked example

1. **Log it (10 minutes, tonight).** Fill in the lapse log and circle the main cause. If it is a competing priority you would choose again, stop and set a pause date.
2. **Shrink it.** Write the version you would do on the worst day of the past month, at most two minutes. It counts as done; if you do more, do not raise the bar.
3. **Re-anchor it.** Write one if-then around a daily event in your current routine: "After I close my laptop at the end of the workday, I put on my shoes and step outside for two minutes."
4. **Set the date.** Tomorrow's cue, or Monday or the 1st if that is three days away or less. Put "Restart day 1" in your calendar.
5. **Remove one obstacle tonight.** Lay out the clothes, open the document, move the charger.
6. **Count days, not a chain.** Track "x of the last 14 days," tiny version included, and read it in your [weekly review](https://learnhubplay.nl/blog/weekly-review-template).

Our suggested ramp, a rule of thumb rather than a finding:

| Weeks | Version | Target |
|---|---|---|
| 1–2 | Tiny version only | 10 of 14 days |
| 3–4 | Tiny on most days, a longer version on up to three | 12 of 14 days |
| 5 onward | Raise length or frequency, never both at once | 12 of 14 days, checked monthly |

If you miss the target, do not push harder: under half of it, shrink the version; between half and the target, move the cue to a more reliable time or place.

**A worked example (hypothetical).** Say you ran three mornings a week until a new job moved your start from 9:00 to 8:15, and your first 45-minute run after a cold left you wrecked. Seven weeks on, the log gives two causes: context changed and habit too big. You restart on Monday, two days away, with the cue "after I close my laptop," a tiny version of shoes on and two minutes outside, and a target of 10 of 14 days. At 12 of 14 by week four, you try 20 minutes on three days.

> **Warning:** Skip the make-up session. We know of no study testing catch-up sessions; our reasoning is that doubling up is likely to rebuild the all-or-nothing pressure the tiny version exists to remove.

## Write the plan for your next miss before it happens

Relapse-prevention treatment prepares for the slip in advance: the 1999 review says clients often get simple written instructions for the moment a lapse happens.

One tested version is small. In Adriaanse and Ten Broeke's Study 2, 209 UK adults who had procrastinated in the past month read a short text about external factors after a failure, then made an if-then plan to reflect on them and keep going. Over the next three to four days the plan group reported about 124 minutes of procrastination against about 229 for controls (d = 0.58). Controls got neither text nor plan, so the plan cannot be separated from the text, and the authors call the evidence tentative: self-reported over days, online participants.

This template follows the same shape, with the cause kept specific and changeable:

```text
IF I miss a day,
THEN within five minutes I write one line naming a cause I can change
     ("the 8:15 start moved my cue"), not a trait ("I'm lazy"),
AND at my next cue I do the tiny version, with no make-up session.

IF I miss two days in a row,
THEN that evening I re-read my lapse log and ask what changed.

IF a full week passes,
THEN I repeat steps 1 to 4 of the re-entry protocol.
```

## Where to start this week

Tonight, fill in the lapse log, write the tiny version and its "after I ___" cue, and set a restart date within three days. Put the miss plan where you will see it, and count days on Sunday.

## Frequently asked questions

### How long a gap is too long to restart the same habit?

No gap is too long to restart; what changes is the plan. The cut-offs are our judgement: we found no study that times how long a habit takes to rebuild after weeks or months away.

### Does my earlier progress count, or am I back at day one?

We found no study of how fast a lapsed habit comes back, so this is our judgement. Wood and colleagues' transfer students kept strong habits when the old context stayed intact, which suggests an unchanged cue helps and a missing one does not. Count days out of the last 14 from your restart date instead of resetting a streak to zero.

### Should I restart the same habit, or replace it?

Replace it when the log points at "no reward you feel," or at a priority you would choose again, and the goal behind the habit still matters: keep the goal, change the behaviour. None of the research above compares restarting with replacing, so that is our judgement.

## Sources and further reading

- [Relapse prevention: An overview of Marlatt's cognitive-behavioral model](https://pmc.ncbi.nlm.nih.gov/articles/PMC6760427) — Larimer, Palmer & Marlatt, Alcohol Research & Health, 1999. Lapse, relapse, lapse management.
- [Relapse dynamics during smoking cessation: recurrent abstinence violation effects and lapse-relapse progression](https://pmc.ncbi.nlm.nih.gov/articles/PMC3296289) — Kirchner, Shiffman & Wileyto, Journal of Abnormal Psychology, 2012. 203 smokers, 1,001 lapses.
- [Beyond prevention: Regulating responses to self-regulation failure to avoid a set-back effect](https://pmc.ncbi.nlm.nih.gov/articles/PMC9291624) — Adriaanse & Ten Broeke, Applied Psychology: Health and Well-Being, 2022. Two if-then plan experiments.
- [How are habits formed: Modelling habit formation in the real world](https://doi.org/10.1002/ejsp.674) — Lally, van Jaarsveld, Potts & Wardle, European Journal of Social Psychology, 2010. Missed opportunities, Armitage contrast.
- [Self-compassion increases self-improvement motivation](https://doi.org/10.1177/0146167212445599) — Breines & Chen, Personality and Social Psychology Bulletin, 2012. Four experiments.
- [The Fresh Start Effect: Temporal Landmarks Motivate Aspirational Behavior](https://faculty.wharton.upenn.edu/wp-content/uploads/2014/06/Dai_Fresh_Start_2014_Mgmt_Sci.pdf) — Dai, Milkman & Riis, Management Science, 2014. Three archival studies.
- [Psychological effects of temporal landmarks: A meta-analysis of affective, cognitive, and self-regulatory processes](https://doi.org/10.1016/j.actpsy.2026.107747) — Wang, Sheng & Zhu, Acta Psychologica, 2026. Pooled effect.
- [Changing circumstances, disrupting habits](https://dornsife.usc.edu/wendy-wood/wp-content/uploads/sites/183/2023/10/Wood.Tam_.GuerreroWitt.2005_Changing_circumstances_disrupting_habits.pdf) — Wood, Tam & Witt, Journal of Personality and Social Psychology, 2005. Habits and context change.