Streaks and daily habits: what the data says

What is a streak, and why does it work?

A streak counts the consecutive days someone performs an action: playing, learning, reading. The counter makes regularity visible and gives it value: the longer the streak, the harder it is to let it go.

In a Journal of Consumer Research paper summarised by the University of Colorado Boulder in April 2023, researchers Silverman and Barasch define a streak as a behaviour performed three or more times in a row. Across seven studies, they show that an intact, tracked streak increases subsequent engagement compared with a broken one. The effect of a break gets worse when people blame themselves for it, and weaker when they can "repair" the streak.

In other words, a streak does not create motivation from scratch. It protects existing motivation by making it costly to stop, and how you design the break matters as much as the counter itself.

How long does it take for a habit to form?

The reference study is by Phillippa Lally and colleagues at University College London, published in 2009 in the European Journal of Social Psychology. 96 volunteers chose an eating, drinking or activity behaviour to carry out daily in the same context (for example "after breakfast"), repeated it for 12 weeks and rated how automatic it felt each day.

The average hides a wide spread: the time to reach 95% of the automaticity plateau ranged from 18 to 254 days. Three lessons matter for a publisher:

• context makes the habit: automaticity comes from linking a cue (morning coffee, opening the app) to an action;
• one slip does not ruin it: in the study, missing one opportunity did not materially affect habit formation;
• inconsistency does: the least consistent participants did not form a habit.

For a daily game, that argues for a stable appointment (same release time, same spot on the site) and for a streak that tolerates the occasional miss rather than punishing every one. It also argues for patience: if a habit takes weeks or months to settle, judging a new game after its first week tells you about curiosity, not about habit. Look at return rates over at least a month before deciding whether a game works for your audience.

What has Duolingo learned from its streaks?

Duolingo has documented the effect of streaks more publicly than anyone. Its figures come from the company itself: they describe its own users, not a general law.

In November 2020, the product team wrote on the Duolingo blog that learners who reach a 7-day streak are 2.4 times more likely to continue using Duolingo the next day. A year earlier, the company had separated the streak from the daily goal, so a single lesson a day was enough to extend it. That change increased the number of learners on a 7+ day streak by over 40%; in the original test, it lifted Day 14 retention by 3.3%, in relative terms. Just over half of daily learners then had a streak of at least 7 days.

A second post, in January 2022, describes smaller tweaks (Duolingo, January 2022). An animation celebrating streak milestones increased the likelihood that a brand-new learner was still active 7 days later by 1.7%. Letting learners equip up to two Streak Freezes at a time increased daily active learners by 0.38%. The post also says learners who reach a 7-day streak are 3.6 times more likely to complete their course.

Taken one by one, these gains look small. Their value comes from scale and accumulation: every detail (celebrating a milestone, forgiving a miss, playing with a friend) adds a little retention.

Do streaks work for publishers' games?

Publishers see the same mechanism. At The Times in London, the head of subscriber retention reported in May 2025 a 9 percentage point retention gap between subscribers who play once a week and those who play two days in a row (Press Gazette, May 2025). It is a correlation: the most loyal readers also play more often.

Streaks also become talking points. In its 2025 round-up, The New York Times said the Wordle of 14 May, whose answer was BONGO, broke 36,337 streaks on its own (TODAY, December 2025). The fact that the paper tracks and publishes that number shows how central streaks are to its players' experience.

Sharing feeds the habit too. In the same piece, Zoe Bell, executive producer of games at The New York Times, mentions an average of 46 Wordle scores shared per second. A shared result reminds everyone who sees it about the daily appointment, and every share reinforces the sharer's own routine. For a publisher, that is a free acquisition channel built directly on players' habits.

How do you design a streak that retains without frustrating?

The evidence above points to a few design principles:

The account link deserves attention. The New York Times made account creation a goal, partly so players could track their stats and streaks (Press Gazette, March 2023). The full case is covered in NYT Games and the subscription bundle. The streak becomes a reason to sign up, and the sign-up becomes first-party data for the publisher.

How SIWA applies these principles

In SIWA's games, which are in French, streaks are on by default. Players see their overall streak, counting consecutive days with at least one game played, and a streak per game, counting consecutive days with a win, along with their best streak and a calendar of the days they played. The overall streak stays alive as long as the player played yesterday, which leaves the whole day to play.

Two social mechanics can be added, off by default and enabled on request: private or public clubs joined with an invite code, with daily and all-time leaderboards, and weekly leagues with divisions, where the top players move up and the bottom ones move down each week. Game selection is on the pro page, and the case for the daily appointment is in the resource on audience retention.