The fresh start effect
Temporal landmarks make people feel disconnected from past failures and willing to try again.
Dates that mark a boundary, such as a new year, a birthday or the start of a week, prompt people to pursue goals at higher rates. The proposed mechanism is that the landmark separates a new self from an old one, which makes past lapses feel like someone else's. The effect is about initiation, not persistence.
How it shows up in software
It explains the annual fitness signup wave and why weekly goals reset on Monday rather than at an arbitrary point. In software the useful move is to make landmarks available on demand, so a user who lapses on Thursday does not have to wait until January to have a clean slate.
Using it well
- Reset goal periods on a boundary the user recognises, such as a week or a month, rather than on their signup anniversary.
- Offer a manual restart control so a lapsed user can draw their own line without losing their account history.
- Time re-engagement around genuine landmarks in the user's own calendar, not around your quarter end.
- Keep past data visible but clearly on the other side of the line, so a restart does not require deleting history.
Where it turns manipulative
- January renewal pricing and landmark-timed upsells exploit a documented moment of optimism to sell contracts people later regret and often cannot cancel.
- Engineering false landmarks, such as invented milestones or arbitrary season resets, trains users to discount the real ones.
- Wiping a user's accumulated progress to manufacture a fresh start takes something they earned in order to restart a counter you own.
Where you have seen it
Apple Fitness
Activity rings reset at midnight and the move goal can be changed weekly, giving a daily and a weekly boundary.
Strava
Monthly and annual challenges open on calendar boundaries rather than on join date.
What the research says
- Dai, Milkman and Riis, 2014Mixed evidence
Found spikes in gym visits, diet-related search queries and goal commitments following new weeks, months, years, semesters and birthdays.
Mostly archival and correlational. Weekly and annual cycles have many confounds, such as institutional calendars and marketing, that compete with the psychological account.
- Dai, Milkman and Riis, 2015Mixed evidence
A field experiment using landmark-framed emails to promote saving found effects that were positive but modest.
Subsequent field tests have produced inconsistent results, including nulls. Treat this as a real but small and fragile effect, not a reliable lever.
Grades are a judgement about the evidence, not about how useful the idea is. Plenty of contested effects are still worth knowing, as long as you do not cite them as settled.