Guide
New Year's resolutions that last past February
Resolutions do not usually fail because people lack willpower. They fail because they were never designed: a big, vague intention, started on a day chosen for its symbolism, with no answer to the question of what happens on the first bad day.
Want this set up as a daily plan, starting 1 January? Two minutes, free, no account.
First, the statistic everyone quotes
You will have read that 80% of resolutions fail by February, or by the second Friday in January. Neither figure has a traceable source. The study that actually followed people through their resolutions — Norcross, Mrykalo and Blagys, 2002 — found that roughly 46% of resolvers were still succeeding six months later, against around 4% of people who wanted to make the same change but did not formally resolve to.
So the honest starting point is the opposite of the cliché: resolving is worth doing. A resolution is a deliberate decision, and deliberate decisions beat vague wishes. The question is how to give yours the best odds.
New Year really is a good moment — but not a magic one
The fresh start effect is real. Dai, Milkman and Riis found that searches for "diet", gym attendance and commitments to goals all rise after temporal landmarks: the new year, the start of a week or month, a birthday. A landmark creates a psychological line between the old you and the new one, and that makes starting easier.
The catch is that the motivation from a landmark fades within weeks. It gets you through the door; it does not keep you there. What keeps you there is a design that still works on a wet Tuesday in February when the motivation has gone. That is what the rest of this page is about.
Set it up in December, start on the 1st
Use the landmark for what it is good at — starting — and do the designing beforehand. Decide the goals, the anchors and the weekly review in the last fortnight of December, so that 1 January is day one of a plan that already exists rather than the day you start thinking about it.
Pick one to three. Not seven.
Every new behaviour draws on the same limited supply of attention in its first weeks, which is exactly when it is most fragile. Seven resolutions means seven things competing for that supply on the same days. Pick the one to three that would change the most, and let the others wait for the next landmark — a new month is only four weeks away.
Start something, rather than stop something
"Stop snacking in the evening" gives you nothing to do; it only gives you something to fail at. "After dinner, make a cup of tea and go for a ten-minute walk" gives you an action. In a year-long study of resolutions, Oscarsson and colleagues (2020) found that approach goals — doing more of something — succeeded more often than avoidance goals. Where you can, rewrite a "stop" as a "start".
Anchor it to something you already do
The most reliable single technique in this whole area is the implementation intention: a plan in the form "when X happens, I will do Y". Across a meta-analysis of 94 studies, Gollwitzer and Sheeran found a medium-to-large effect on actually following through.
The trick is choosing X well. It should be something that already happens every day without effort — putting the kettle on, sitting down on the train, brushing your teeth — so the new behaviour gets a cue it did not have to create. "I'll read more" becomes "when I get into bed, I read ten pages before I pick up my phone."
Make it small enough to do on a bad day
Plan for your worst reasonable day, not your best one. Ten push-ups, one page, a five-minute tidy. A small version that you do on 25 days in January builds more habit than an ambitious version you do on 8. You can scale up in February once the cue is doing its job; you cannot build a habit out of days you did not show up for.
Count the days you showed up — not the streak
Streak-counting apps have a failure mode: the first missed day resets the number to zero, and zero feels like permission to stop. Research on dieting found the same pattern — people who break a rigid rule often abandon it entirely, sometimes called the "what-the-hell effect" (Polivy and Herman).
A better measure is how many days out of the last seven or thirty you showed up. Five out of seven is a good week. It stays a good week even though the streak broke. Flexible consistency beats rigid streaks — and it is far more honest about what a habit actually looks like.
Decide now what happens after a miss
You will miss a day. In the best-known study of habit formation, Lally and colleagues (2010), a single missed day did not materially affect how quickly the habit formed. What does the damage is the story you tell about the miss.
So write the rule down before you need it: never miss twice, and never compensate. Do not try to do double tomorrow. Just do tomorrow's.
How long it actually takes
The "21 days" figure is a myth, traced back to an observation about surgery patients adjusting to their appearance. In Lally's study the median time to automaticity was 66 days, with an enormous range — 18 to 254 — depending on the person and the behaviour. Something simple, like a glass of water with breakfast, got there fast; something effortful, like fifty sit-ups, took much longer.
For a resolution started on Friday 1 January 2027, the median lands around Monday 8 March. That is worth knowing in advance, because it tells you where the hard part is: January, and most of February. If it still feels like effort in the middle of February, that is not failure. That is the median.
A ten-minute weekly review
Every Sunday: which days did it happen, and on the days it did not, what got in the way? If the same thing broke it three times, the anchor is wrong — change the cue, not your opinion of yourself. This is the single cheapest thing that turns a resolution into a plan, and it is the step almost everyone skips.
Plan my resolutions →or build it with Claude or ChatGPT
Questions
Is the 80%-fail-by-February figure real? Not as far as anyone can trace. The study that followed resolvers found around 46% still succeeding at six months.
How many resolutions? One to three. More than that and they compete for the same attention in the weeks when they are most fragile.
Should I wait for 1 January? The date helps you start. Design the plan in December so the landmark is spent on starting, not planning.
How long until it is a habit? A median of 66 days in the best-known study, with a range of 18 to 254. Early March, for a 1 January start.
What if I miss a day? Resume the next day and do not make up for it. One miss does not reset the habit; abandoning it does.