Many New Year's resolutions last well past January, and the widely shared "92% fail" figure is not a research result. It comes from a survey done for a 2006 self-help book, which found that only 8% of people always achieve their resolutions. Studies that actually followed resolvers found that about 40% were still on track after six months, and 55% of the people who answered the one-year follow-up in a 2020 study of 1,066 considered themselves successful.
That does not mean resolutions are easy. The same research shows a steady drop-off over time, and it points to a few habits that predict who keeps going.
The Claim: 92%, 80% and "Quitter's Day"
Three statistics dominate every January: that 92% of resolutions fail, that about 80% are dropped by February, and that the second Friday of January is "Quitter's Day," when most people give up. In 2027, the second Friday of January falls on January 8.
All three get repeated as if they were the verdict of one big study. They are not, and they measure three different things.
Where Each Number Really Comes From
The "92%" is the flip side of an "8%" figure. Stephen Shapiro, the author of the 2006 book Goal-Free Living, reports that research done with Opinion Research Corporation found "only 8% of people are always successful in achieving their resolutions." Flip 8% and you get 92%, but the survey counted people, not resolutions. Someone who kept four resolutions and broke one fell outside the 8%.
The same survey reported that 19% succeed every other year, 49% have infrequent success and 24% never succeed. Some articles later credited the 8% to a University of Scranton study in the Journal of Clinical Psychology. The Scranton team's own papers report very different numbers, shown below.
The "80% fail by February" figure circulates in news articles as "by some estimates," and no study behind it has been found. The trail ends at other web pages, not at published research.
"Quitter's Day" is real data, but narrower. On January 12, 2018, the second Friday of the month, the fitness app Strava said its analysis of 31.5 million activities uploaded in earlier Januaries showed that this was when exercise uploads dropped off most. By 2020, Strava was calling its predicted date, January 19, "Quitter's Day." The date has changed from year to year. The measure is a dip in uploads to one app on one day, not a count of people who quit.
| Claim | What is behind it | Supported? |
|---|---|---|
| 92% of resolutions fail | Flip of "8% always succeed" from a survey for a 2006 self-help book; counts people, not resolutions | No |
| 80% fail by February | "By some estimates" in news articles; no study found | No |
| Quitter's Day is the second Friday of January | Strava's 2018 analysis of 31.5 million uploads; later dates differ by year | Partly: a dip in app activity, not proof that most people quit |
What Studies Actually Show
The best-known follow-up studies come from psychologist John Norcross and colleagues at the University of Scranton. In a study published in 1988, Norcross and Vangarelli followed 200 resolvers for two years. Of them, 77% were still keeping their resolutions after one week, 55% after one month, 43% after three months, 40% after six months and 19% after two years. Even the successful resolvers stumbled: 53% of them had at least one slip, and the average was 14 slips over the two years.
In 2002, Norcross, Mrykalo and Blagys followed 159 resolvers and 123 people who wanted to change a problem but did not make a resolution. At six months, 46% of the resolvers had been continuously successful, compared with 4% of the non-resolvers. Confidence, skills for change and readiness to change, measured before January 1, predicted who succeeded. These studies relied on people reporting their own progress.
A 2020 study in PLOS ONE by Oscarsson and colleagues tracked 1,066 members of the public who signed up after Swedish media and social media coverage in late December 2016. After one year, 55% of the people who responded considered themselves successful. The authors note that volunteers were probably more motivated than average, so the figure may be higher than for the general public.
The most popular resolutions in that study were about physical health, weight and eating. Goals framed as something to start (approach goals) were kept more often than goals framed as something to stop (avoidance goals): 58.9% versus 47.1% successful at one year.
How Long a New Habit Takes
A 2010 study by Lally and colleagues followed 96 volunteers who picked one eating, drinking or exercise behavior to repeat daily for 12 weeks; 82 gave enough data to analyze. The time people took to reach a plateau of automatic behavior ranged from 18 days to an estimated 254 days. The famous 66 days is the typical (median) time among the 39 people whose data fit the researchers' model well, not a deadline that applies to everyone.
The longest estimate lies beyond the 84-day study window, so it is a projection. Missing a single day did not noticeably hurt the process.
The Verdict
Verdict: Partly true. The numbers behind "92% fail" and "80% fail by February" are not supported, and studies find that roughly 40% to 55% of resolvers report success at six months to a year. But keeping a resolution does get harder over time, and in the 1988 study only 19% were still on track after two years.
What Helps, According to Research
These are findings from studies, not personal advice, and results vary between people.
- Plan the when and where. Gollwitzer and Sheeran's 2006 review of 94 tests found that "if situation Y, then I will do X" plans had a medium-to-large positive effect on reaching goals. A much larger 2024 review by the same research group, covering 642 tests, found the benefit ranged from small to fairly large depending on what was measured, and was bigger when plans used a clear if-then format, people were highly motivated and they rehearsed the plan. A 2002 study of 59 students found no difference after one month, so the effect is not guaranteed.
- Frame goals as something to start. In the 2020 study, goals framed as something to do (such as walking more) were kept more often than goals framed as something to stop (such as snacking less).
- Track progress, ideally in writing. A 2016 review by Harkin and colleagues of 138 studies with 19,951 participants found that monitoring progress improved goal attainment, and more so when progress was physically recorded or reported to others.
- Get some support, but keep it simple. In the 2020 study, people who named a supporter and received a little goal-setting information did slightly better than a control group. A third group that also named a supporter but got extra exercises and instructions did no better than the controls, so more help was not automatically better.
- Expect a slow start and forgive a miss. Lally's study found habits take weeks to months and one missed day did little harm, and in the 1988 study most successful resolvers had slipped at least once.





