Blog · · Updated · 8 min read

Do rewards build habits? At first, yes

In a study of more than 61,000 gym members, the best of 53 ideas for getting people to exercise was a nine-cent bonus for coming back after a missed workout. Rewards work, above all at the start. What makes a habit last is something else: repeating the action in the same situation, until it runs without the reward.

What counts as a reward

Psychologists call it positive reinforcement: anything pleasant that follows an action and makes it more likely to happen again. Praise for tidying up, a treat for a dog that sits. In an app the rewards are small: a box that turns green, a sound, a ring that fills, a badge. A streak counter is both kinds at once. It rewards you every day it grows, and takes something away the day it drops to zero.

The brain keeps track of rewards in a particular way. In monkeys, nerve cells that release dopamine, a chemical tied to reward, fired when a reward arrived by surprise. Once a cue reliably announced the reward, they fired at the cue instead, and when an expected reward failed to come, their activity dipped.1 With time, the cue carries the signal, and an expected reward that goes missing feels like a loss.

Rewards start habits. They don't keep them going

Rats trained briefly to press a lever for sugar stopped pressing once the sugar had been made unpleasant to them. Rats trained for longer kept pressing for a reward they no longer wanted.2 That is now the standard test of a habit: the action continues after its reward has lost its value.

People do the same. At a cinema, viewers with a strong popcorn habit ate about as much week-old popcorn as fresh, though they rated it poorly. In a meeting room, or eating with their other hand, they ate less of the stale kind.3 The habit belonged to the setting and the routine, not to the taste.

Rewards do their work early. A review of habit research describes the brain's reward signal strengthening the link between a situation and an action while the action is new, and fading once the reward is expected. Habits also form more readily when rewards are only loosely tied to each action, and more slowly when people weigh up the payoff each time.4 Rewards help you do the repetitions; the repetitions build the habit.

A reward now beats a reward later

Most habits worth having pay off late. Yet in five studies, what predicted whether people kept exercising or studying, within a session or over months of New Year’s resolutions, was how much they enjoyed it at the time. The long-term benefits they said they were after did not predict it.5

An app cannot make a run enjoyable, but it can add a small, certain signal, the moment an action is done, that it counted. That is why ticking a goal in Allez! plays a short sound.

Praise helps. Payment can backfire

In a classic experiment, preschoolers who liked drawing were promised a “Good Player” certificate for it, given one as a surprise, or given nothing. Later, in free play, the children who had drawn for the promised certificate spent about half as much time drawing as the others.6

A review of 128 experiments found the same pattern again and again. When people worked for a reward they had been promised, money or a certificate, they later spent less time on an activity they had enjoyed. Unexpected rewards changed nothing. Praise did the opposite: people spent more time on the activity and found it more interesting. Wording mattered: “you should keep up the good work” motivated less than simply saying how well someone had done. And for dull tasks, rewards did no reliable harm, since there was little enjoyment to crowd out.7

The useful kind of reward tells you what you did (“3 of 5 done”) rather than paying you for it. It is the same line, between doing something because you value it and doing it for a payoff, that runs through the research on motivation in How long does it take to build a habit?

Badges, points and leaderboards: a mixed record

Adding game elements to things that are not games, known as gamification, works on average. In a review of 16 randomised trials, gamified programmes got people moving more, though about 14 weeks after they ended the effect was a third of its size.8

The details are less flattering. On Stack Overflow, a question-and-answer site for programmers, people did markedly more of whatever a badge rewarded in the days before they earned it, then went back almost at once to their usual level.9 A reanalysis suggests only about one user in five responded to badges at all.10 In a semester-long university course with badges and a leaderboard, students ended up less motivated, and with lower exam scores, than in the same course without them.11

One effect is reliable: visible progress pulls people towards a reward. At a car wash, a loyalty card that came with a head start was completed far more often.12

Car-wash customers who earned their free wash
A card with 8 empty spaces A card with 8 empty spaces: 19% 19% A card with 10 spaces, 2 already stamped A card with 10 spaces, 2 already stamped: 34% 34%
Both cards needed eight more washes for a free one. The only difference was a head start that cost the car wash nothing.

Streaks: a reward that can be taken away

A streak motivates while it lasts. People shown an unbroken record of their own behaviour do more of it than people shown the same record with a break, and a break hurts most when people blame themselves for it.13 There is more on this, and on the “what-the-hell effect” that can follow a lapse, in How long does it take to build a habit? One more finding belongs here: in a study of a popular habit app, streaks and reminders helped people repeat new behaviours but left them depending on the app. When they stopped using it, the behaviour often stopped too.14

What helps after a miss? The gym study, a “megastudy” of 61,293 members, tested 53 four-week programmes at once. The basic one, with planning, reminders and about $0.22 in points per visit, raised visits by 9%. The best added about $0.09 for coming back after a missed workout, and raised them by 27%.15 The money was trivial. What it rewarded was the return, and Allez! gives a prize for exactly that: a good day straight after a missed one.

Gym visits, compared with the control group
Planning, reminders and $0.22 a visit Planning, reminders and $0.22 a visit: +9% +9% The same, plus $0.09 for coming back after a miss The same, plus $0.09 for coming back after a miss: +27% +27%
Members otherwise went 1.48 times a week. As the best of 53, the top result is probably somewhat inflated, and only 8% of the programmes still had an effect in the four weeks after they ended.

Surprise rewards: powerful, and not what you want

Rewards that come at random, like a slot machine’s, are famously good at keeping people coming back. In more than a million social media posts, people posted the way animals learn where food is: more often after many likes, less often after few.16 Surprise bonuses and mystery boxes keep people in an app. They do not obviously build a habit outside it, and they make the app, not the behaviour, what you return to.

What makes a habit automatic

One of the few experiments to test positive reinforcement inside a habit app compared ways of supporting a new daily habit over four weeks. Reminders helped people repeat it but kept it from becoming automatic. Linking it to an event in their day made it more automatic. Positive reinforcement made no difference.17 A 2024 review of 41 digital habit programmes found that 16 used virtual rewards, but only 2 tested whether the rewards did anything.18 What builds automaticity is a stable cue; there is more on choosing one in How long does it take to build a habit?

If you are building a habit

What this research cannot tell you

Common questions

Is positive reinforcement the same as gamification?
No. Positive reinforcement is anything pleasant after an action that makes the action more likely. Gamification borrows game elements such as points, badges and leaderboards. Some of them reinforce; some, like a leaderboard you are sliding down, can do the opposite.

Do streaks work?
While they last, yes: an unbroken record makes people more likely to continue. The trouble comes the day one breaks, especially if you blame yourself for it.13

Should I promise myself a big reward at the end of the month?
It is unlikely to help at the moment that matters, when you are deciding whether to go. In five studies, what predicted persistence was enjoyment during the activity, not the benefits that come later.5 Something pleasant during the action, or straight after it, is the better bet.

When can I stop rewarding myself?
No study gives a date. A sign is when the cue alone gets you started, without weighing it up. By then the link between the situation and the action carries the habit, and the reward matters less.4

The studies behind this

  1. Schultz, Dayan & Montague (1997). A neural substrate of prediction and reward. Science.
  2. Dickinson (1985). Actions and habits: The development of behavioural autonomy. Philosophical Transactions of the Royal Society B.
  3. Neal, Wood, Wu & Kurlander (2011). The pull of the past: When do habits persist despite conflict with motives? Personality and Social Psychology Bulletin.
  4. Wood & Rünger (2016). Psychology of habit. Annual Review of Psychology.
  5. Woolley & Fishbach (2017). Immediate rewards predict adherence to long-term goals. Personality and Social Psychology Bulletin.
  6. Lepper, Greene & Nisbett (1973). Undermining children’s intrinsic interest with extrinsic reward: A test of the “overjustification” hypothesis. Journal of Personality and Social Psychology.
  7. Deci, Koestner & Ryan (1999). A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin.
  8. Mazeas, Duclos, Pereira & Chalabaev (2022). Evaluating the effectiveness of gamification on physical activity: Systematic review and meta-analysis of randomized controlled trials. Journal of Medical Internet Research.
  9. Anderson, Huttenlocher, Kleinberg & Leskovec (2013). Steering user behavior with badges. Proceedings of the 22nd International Conference on World Wide Web.
  10. Hoernle, Kehne, Procaccia & Gal (2020). The phantom steering effect in Q&A websites. arXiv preprint.
  11. Hanus & Fox (2015). Assessing the effects of gamification in the classroom: A longitudinal study on intrinsic motivation, social comparison, satisfaction, effort, and academic performance. Computers & Education.
  12. Nunes & Drèze (2006). The endowed progress effect: How artificial advancement increases effort. Journal of Consumer Research.
  13. Silverman & Barasch (2023). On or off track: How (broken) streaks affect consumer decisions. Journal of Consumer Research.
  14. Renfree, Harrison, Marshall, Stawarz & Cox (2016). Don’t kick the habit: The role of dependency in habit formation apps. CHI Conference Extended Abstracts on Human Factors in Computing Systems.
  15. Milkman and colleagues (2021). Megastudies improve the impact of applied behavioural science. Nature.
  16. Lindström, Bellander, Schultner, Chang, Tobler & Amodio (2021). A computational reward learning account of social media engagement. Nature Communications.
  17. Stawarz, Cox & Blandford (2015). Beyond self-tracking and reminders: Designing smartphone apps that support habit formation. CHI Conference on Human Factors in Computing Systems.
  18. Zhu, Long, Wang, Lee, Zhang & Wang (2024). Digital behavior change intervention designs for habit formation: Systematic review. Journal of Medical Internet Research.