Why what worked for someone else doesn’t always work for you

Someone tells you the thing that fixed their sleep. You try it. It doesn’t take. Here is what the research says about why, and why willpower isn’t the whole story.

Crowdseek · 14 September 2026 · about 5 minutes to read. Every study is linked at the end, with what it does and doesn’t show.

It usually arrives in good faith. A friend stopped drinking after eight and started sleeping through. A colleague swears by the 6am walk. Your brother-in-law has a whole system.

So you try it. And for a week or two it half works, or it doesn’t, and then a bad Tuesday comes along and it quietly stops. You don’t decide to abandon it. It just isn’t there any more.

Most of us read that as a personal failing. The research doesn’t.

First, the thing everyone has been told is wrong

You have probably heard it takes 21 days to form a habit. It doesn’t, and the number never came from habit research at all. It traces to a 1960 book by a plastic surgeon, describing how long his patients took to get used to their changed appearance.³

When researchers at University College London actually measured it, they asked volunteers to repeat one new daily action in the same setting and tracked how automatic it felt. The median time to reach the top of that curve was 66 days. The range ran from 18 days to 254.¹

The useful finding isn’t 66. It’s the spread.

Estimated habit-formation times in that study varied from 18 to 254 days. A later review of twenty studies found much the same picture, and an even wider individual range.² So if something hasn’t stuck in three weeks, nothing has gone wrong. You were told a number that was never true.

Wanting it isn’t the problem

The obvious explanation for a thing not sticking is that you didn’t want it enough. That is measurable, and it has been measured.

Pooling 25 samples and nearly 30,000 people, researchers found that among people who genuinely intended to be physically active, fewer than half followed through. The gap between intending and doing was 47.6%.⁵ An older analysis of experiments found that when you successfully change what someone intends to do by a lot, you change what they actually do by rather less.⁴

Intending is not nothing. It is just a long way short of enough, for almost everybody.

The study that explains it best moved 115 people to a new university

In 2005, researchers followed students transferring between universities. They measured three ordinary habits before the move and again a month after: exercising, watching television, and reading the paper.⁶

Strong habits survived the move. But only under one condition.

Exercise habits survived when people kept exercising in the same kind of place. Change the setting, and the habit went. Television habits survived when the new viewing situation felt like the old one. Newspaper reading survived when people still read it with other people, as they had before.

Same people, similar intentions. When important parts of the context changed, established habits were less likely to persist.

The action is what travels between people. The context that made it work is what stays behind.

That is the whole of it. When your friend tells you what worked, they are handing you the visible part. What they cannot hand you is the hour they get home, the person who is or isn’t in the kitchen, how they slept, what else is already in the week, and the twenty other things that made the action make sense for them.

We copy people anyway, and we don’t notice

Copying isn’t a flaw. It is one of the most reliable findings in the field. A review of 69 experiments, across more than 5,800 people and forty years, found that people take their cue from those around them for what and how much to eat. 64 of the 69 found the effect.⁷

Two details matter. We copy people we want to be like, or think we resemble. And it happens largely without our noticing.

So the instinct to borrow someone else’s answer is deeply human. It is also why the answer so often arrives with nothing attached to it.

Stopping is common. The numbers don’t tell us why.

A study of nearly 390,000 fitness app users tracked what happened to beginners. Around 9 in 10 stopped within a year, and the median time to dropping out was 19 weeks.⁸ A separate analysis of mental health and wellbeing apps found that of the people who install one, around 3 to 4% are still using it a month later.⁹

Those numbers show the scale of the problem. They do not explain it. Someone who solved their problem and stopped looks exactly like someone who gave up, and leaving an app is not the same as stopping exercise. The studies measure how hard persistence is, not why it fails.

Our own reading, and it is a reading rather than a finding, is that copy, try, struggle, start again is the only method most of us were ever given. If context is what holds a behaviour in place, then keeping the context and adjusting inside it should be a more useful response than starting over. That is what we have built. It is not what these studies prove.

The genuinely hopeful part

If context is what holds a behaviour in place, then a change of context is an opportunity, not just a disruption.

Researchers call this the habit discontinuity hypothesis. In a field experiment with 800 people, a behaviour-change intervention worked better on people who had recently moved house, and the window appeared to stay open for about three months.¹⁰ A study of 18,053 people in a UK national survey found the same pattern, fading as time since the move increased.¹¹

And one of the better-studied ways to help an intention become action is to decide in advance exactly when and where it happens. The classic review of 94 studies found a meaningful effect on whether people actually do the thing.¹² It is twenty years old now, and it only helps people who already want the outcome. Within those limits it holds up.

What we do instead

Crowdseek starts at the other end. Not with an action that worked for somebody else, but with what you actually want back, and the life the plan has to fit inside.

You get a small plan, not an overhaul. Each morning we ask one question and you answer in a line. When something stops fitting, we change the smallest useful thing rather than handing you a new programme. There is no score, and a hard week leaves no backlog to make up.

Right now it does sleep. Movement opens next.

Which one sounds like you?

Related reading: why sleep trackers can make tired people feel worse. Sleep clinicians have described a pattern they call orthosomnia, where chasing a better score makes sleep worse.¹³ It is the same argument as this article, applied to one device.