Does Habit Stacking Actually Work?

TL;DR

Habit stacking has been put to the test head-to-head, once. In a preregistered randomized trial, anchoring a new behaviour to an existing routine worked — and worked no better than anchoring it to a time of day.
The if-then planning literature it borrows from is genuine but smaller than advertised: 0.36 across 642 tests, and 0.15 once publication bias is corrected for.
We read the 15 pages ranking for this question. Twelve call the method science-backed. None cites a trial that compared it with any other kind of cue.
What predicts success is not the anchor. It is anchor stability, the order of the two actions, and how many plans compete in the same part of your life.
Median time to automaticity in that trial was 59 days, with individual values from 4 to 335. Fewer than one in four participants reached it at all.

Habit stacking is one sentence: After I [current habit], I will [new habit]. Pour the coffee, then meditate for a minute. Brush your teeth, then floss. The old routine does the remembering, so you do not have to.

It is the most repeated piece of habit advice of the past decade, and the claim attached to it is usually some version of “grounded in decades of behavioural science.” That claim is where this article starts, because it is checkable.

What the research actually tested

One trial put the anchor against a clock

In 2021, Keller and colleagues ran the comparison directly, and preregistered it before collecting data. They randomized 192 adults into two planning conditions. One group tied a new daily behaviour to an existing routine — after breakfast. The other tied the same behaviour to a time — at 9 a.m. Everything else was held constant across 84 days of daily reports.

Both groups built automaticity. Neither built more of it than the other. The difference between the routine anchor and the clock did not reach significance on any measure the study was designed to detect it on.

What did predict automaticity was repetition: how many days the person actually performed the plan. The anchor was the packaging, not the active ingredient.

Two numbers from that trial deserve to travel further than they have. Median time to the automaticity plateau was 59 days, with individual values spread from 4 to 335. And of 117 participants with usable curves, 27 met the plateau criterion at all — under a quarter.

Two smaller studies point the same way. Stawarz, Cox and Blandford ran a four-week experiment with 96 completers comparing an event-based cue (lunch), a time-based reminder, and no cue. The event cue produced more automaticity than reminders — but the gap between having a cue and having none was not significant, and the reminder group was better at actually doing the behaviour. Stecher and colleagues, in a four-arm trial of 161 students, found that habit strength did not differ between anchoring and non-anchoring conditions at any measurement point.

What we counted in the search results

On 27 July 2026 we read every page we could open that ranks for does habit stacking work, is habit stacking evidence based and habit stacking science. Fifteen pages opened, and all fifteen were read in the body text rather than the summary or the related-articles rail.

12 of 15	describe the method as backed by science or research
4 of 15	name a single study by author and year; the rest use “research shows” or “researchers at Stanford”
0 of 15	mention any trial comparing stacking with another kind of cue
9 of 15	discuss conditions under which the anchor fails; four of them in any depth

Two claims in that set do not survive a check. One page credits “researchers at Stanford” with an 80% gain in adherence and links to a behaviour diagram rather than a study. A second claim circulating well beyond the top results — a 2025 Journal of Applied Psychology paper reporting a 64% improvement from habit stacking — has no traceable source at all.

The limits of our count are worth stating. One search tool, United States results, one day, one location. Personalized results will differ, and the ranking order will have moved by the time you read this. The counts describe what those fifteen pages say, not what the whole internet says.

The borrowed evidence is real, and smaller than advertised

Habit stacking is a popular repackaging of implementation intentions — Peter Gollwitzer’s if-then planning, formalised in the 1990s. The evidence usually cited for stacking is really the evidence for that.

The number that travels is d = 0.65, from a 2006 review of 94 studies by Gollwitzer and Sheeran. It dates from 2006, and its own authors have since revised it. In 2025, Sheeran, Listrom and Gollwitzer pooled 642 tests and reported d = 0.36. Egger’s regression in that paper returns a value the authors themselves call substantial publication bias. Trim-and-fill leaves the estimate at 0.35; a robust Bayesian model-averaged estimate drops it to 0.15.

The same pattern shows up elsewhere. Webb and Sheeran found that interventions changing intention by d = 0.66 changed behaviour by d = 0.36. In physical activity specifically, Belanger-Gravel and colleagues pooled the physical activity trials and got d = 0.31 at post-test, falling to 0.24 at follow-up.

None of this makes planning useless. A planning effect of 0.15 to 0.36 is a real effect, obtained cheaply, and most self-improvement advice cannot show one at all. It is simply not the effect the slogan implies.

What holds up, and what does not

The number you have seen	What the source says
“It takes 66 days to form a habit”	A median in Lally’s 2010 study, range 18 to 254 days. The curve fitted well for 39 of the 96 people who started
“If-then planning has an effect of 0.65”	The 2006 figure. The 2025 update on 642 tests gives 0.36 raw and 0.15 bias-corrected
“Planning took compliance to 100%”	Orbell’s 1997 study: 64% against 14% across 155 women. The 100% figure is a subgroup of 33 with strong prior intention
“Flossing after brushing builds a stronger habit”	Judah’s 2013 study, 50 participants, allocated alternately rather than randomly. At four weeks the difference did not reach significance

Four documented ways an anchor breaks

This is the part the search results mostly skip, and it is the part that decides whether your stack survives past week two.

1. The anchor is not as fixed as you think. Stacking assumes the old habit happens at a stable time, in a stable place. “After my coffee” is not one cue if the coffee happens at 6.40 on weekdays, at 10.00 on Sundays, and in a different room when you travel. Wood, Tam and Guerrero Witt followed 115 students through a transfer to a new university and watched well-practised habits collapse when the surrounding context changed. If the anchor moves, the thing bolted to it moves with it.

2. You put the new action first. Judah and colleagues tested flossing before brushing against flossing after brushing. The proposed mechanism is clean: an established routine runs on autopilot and skips straight past anything inserted in front of it. The direction of their result favours going second, but the study allocated participants alternately rather than randomly, and at four weeks the difference did not reach significance. Treat “put the new action after” as a sensible default, not a proven law.

3. You stacked three things onto the same part of your life. Verhoeven and colleagues found that multiple simultaneous plans interfered with each other — but only when they targeted the same goal domain. Plans in unrelated domains did not compete. Three new plans around eating will fight; one around eating and one around a morning walk will not.

4. You wrote the plan as a refusal. Plans phrased as then I will not eat the chocolate can backfire, priming the very thing they name. In Adriaanse’s studies, negation plans led to more snacking than a plain goal did, while replacement plans — then I will eat an apple — reduced it. Habit stacking is naturally a replacement format, which is one of its quieter strengths. Keep it that way.

There is a fifth failure mode that the literature on stacking does not address, and it is worth naming: the old habit does not disappear when you build a new one on top of it. That is a separate mechanism, and we cover it in why your brain never deletes a bad habit.

How to build a stack that survives

Pick the anchor by stability, not by convenience. The shift is from “what do I do every day” to “what do I do every day at roughly the same time, in the same place, including weekends.” First move: look at the last seven days honestly and pick the one routine that did not move. If nothing qualifies, a fixed time in the calendar is a legitimate alternative — the trial says it performs the same. Our comparison of time-based systems covers what that looks like in practice.

Put the new action after the anchor, and make the seam physical. The shift is from sequencing in your head to sequencing in the room. First move: leave the object for the new behaviour where the old routine ends — the floss beside the toothbrush, the notebook on the coffee machine. When you travel, the object travels.

Run one new plan per life domain. The shift is from stacking a morning routine to stacking a single behaviour. First move: if you have three plans queued, keep the one whose anchor scored best on stability and postpone the other two by a month.

Write the plan for the day the anchor disappears. Every anchor eventually fails: illness, travel, a new job, a moved desk. The shift is from treating that as a broken streak to treating it as a known event with a prepared response. First move: write a single backup line now — if I am away from home, then I will do it after I put my bag down. Wood’s transfer students lost habits precisely because no such line existed.

Count repetitions, not streaks. Repetition was the predictor in the 2021 trial, and repetition survives a missed day. A streak counter does not, and a broken streak is a common exit point.

Where this stops working

Planning helps people who already want the outcome. It is a bridge across the gap between intention and action, and if the intention is not there, there is nothing to bridge. If you are stacking a behaviour you do not actually want, no anchor will rescue it.

Effortful behaviours are a poor fit for a light anchor. Lally’s study found a median of 66 days to automaticity with a range from 18 to 254, and the slow end was populated by demanding behaviours. Fifty squats after your coffee is not a stack; it is a workout with a coffee-flavoured reminder. Our full account of that timeline is in how long it really takes to form a habit.

Where the problem is avoidance rather than forgetting, stacking is aimed at the wrong target. Forgetting is a cue problem; avoidance is an emotional regulation problem, and the two need different tools — see why you procrastinate, ranked by evidence.

In diagnosed mental health conditions the evidence is promising and wide open: a review by Toli, Webb and Hardy pooled 29 tests and reported d = 0.99, with a confidence interval running from 0.40 to 1.60 and a significant test for publication bias. An interval that wide is a signal to be careful, not confident. If the behaviour you cannot start is sleep, eating, medication or self-care, and the difficulty has persisted for weeks, that is a conversation for a clinician rather than a planning template. Nothing here is medical advice.

The boring bottom line

Habit stacking is a good default and a bad explanation. It works because you repeat a small action in a consistent context, not because attaching it to an existing habit does something a fixed time could not. When it fails, the cause is almost always an anchor that was never as stable as it looked.

Choose a routine that does not move, put the new action after it, run one at a time, and write the line for the day it breaks. That is the whole method, and it is duller than the version that sells.

This article belongs to our habits and discipline collection, where we track which habit advice survives contact with the evidence.

Sources

Keller, J., Kwasnicka, D., Klaiber, P., Sichert, L., Lally, P., & Fleig, L. (2021). Habit formation following routine-based versus time-based cue planning: A randomized controlled trial. British Journal of Health Psychology. doi:10.1111/bjhp.12504
Sheeran, P., Listrom, O., & Gollwitzer, P. M. (2025). The when and how of planning: Meta-analysis of the scope and components of implementation intentions in 642 tests. European Review of Social Psychology, 36(1), 162-194. doi:10.1080/10463283.2024.2334563
Stawarz, K., Cox, A. L., & Blandford, A. (2015). Beyond self-tracking and reminders: Designing smartphone apps that support habit formation. Proceedings of CHI 2015, 2653-2662. doi:10.1145/2702123.2702230
Stecher, C., Chen, C., Codella, J., Cloonan, S., Hendler, J., et al. (2024). Combining anchoring with financial incentives to increase physical activity: A randomized controlled trial among college students. Journal of Behavioral Medicine, 47, 751-769. doi:10.1007/s10865-024-00492-4
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. doi:10.1016/S0065-2601(06)38002-1
Webb, T. L., & Sheeran, P. (2006). Does changing behavioral intentions engender behavior change? A meta-analysis of the experimental evidence. Psychological Bulletin, 132(2), 249-280. doi:10.1037/0033-2909.132.2.249
Belanger-Gravel, A., Godin, G., & Amireault, S. (2013). A meta-analytic review of the effect of implementation intentions on physical activity. Health Psychology Review, 7(1), 23-50. doi:10.1080/17437199.2011.560095
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. doi:10.1002/ejsp.674
Wood, W., Tam, L., & Guerrero Witt, M. (2005). Changing circumstances, disrupting habits. Journal of Personality and Social Psychology, 88(6), 918-933. doi:10.1037/0022-3514.88.6.918
Judah, G., Gardner, B., & Aunger, R. (2013). Forming a flossing habit: An exploratory study of the psychological determinants of habit formation. British Journal of Health Psychology, 18, 338-353. doi:10.1111/j.2044-8287.2012.02086.x
Verhoeven, A. A. C., Adriaanse, M. A., de Ridder, D. T. D., de Vet, E., & Fennis, B. M. (2013). Less is more: The effect of multiple implementation intentions targeting unhealthy snacking habits. European Journal of Social Psychology, 43(5), 344-354. doi:10.1002/ejsp.1963
Adriaanse, M. A., et al. (2011). Planning what not to eat: Ironic effects of implementation intentions negating unhealthy habits. Personality and Social Psychology Bulletin, 37(1), 69-81. doi:10.1177/0146167210390523
Orbell, S., Hodgkins, S., & Sheeran, P. (1997). Implementation intentions and the theory of planned behavior. Personality and Social Psychology Bulletin, 23(9), 945-954. doi:10.1177/0146167297239004
Gollwitzer, P. M., & Brandstatter, V. (1997). Implementation intentions and effective goal pursuit. Journal of Personality and Social Psychology, 73(1), 186-199. doi:10.1037/0022-3514.73.1.186
Toli, A., Webb, T. L., & Hardy, G. E. (2016). Does forming implementation intentions help people with mental health problems to achieve goals? British Journal of Clinical Psychology, 55(1), 69-90. doi:10.1111/bjc.12086

Every source above was checked against the publisher record before publication. Where a figure appears in more than one form in the original, both are reported here.