You probably have one on your phone. A fitness app you downloaded full of enthusiasm: on January 1st, on a Monday, after a photo you didn't like. The first week, you opened it every day. You protected your streak, celebrated your first badge, showed someone your weekly summary. By week three, you were opening it out of obligation. And somewhere between week four and week six, you stopped opening it altogether, without a specific day you could point to.
It wasn't a lack of willpower. It's a curve, and it's been measured.
I'm on the other side of that curve. I'm building Dromo, a running app for Latin America: every run earns points based on how far you went, the city has rewards hidden on the map that you reach by running, and every week you compete in a league against other runners. The question I care about isn't whether people will download it. It's what will happen in week six.
Before designing a single mechanic, I went looking for what the evidence says. I found bad news, a warning, and a way out. All three changed how the app is built.
The summer everyone went out walking
July 2016. Parks filled up with people staring at their phones at midnight. People who wouldn't walk to the corner store were circling their neighborhoods hunting for invisible creatures. Without meaning to, Pokémon GO became the largest public health experiment in history.
And since it happened to millions of people wearing watches and carrying phones, it was all recorded.
Althoff, White, and Horvitz (2016) combined data from 32,000 Microsoft Band users with their web searches. The most engaged players walked 1,473 more steps a day, 26% more than before, for an entire month. No health campaign had ever achieved anything like it.
But Howe and colleagues (2016), in the BMJ, followed young players for longer and saw the other half of the story: 955 more steps a day in the first week, followed by a slow, steady decline. By week six, they were walking exactly as much as they had before installing the game. A later review of seventeen studies, by Khamzina and colleagues (2020), confirmed the pattern.
Six weeks. The same window in which you quit your app.
The bad news: the effect doesn't stick
Pokémon GO could have been a one-off. It wasn't.
In 2022, Mazéas and colleagues pooled every controlled trial that had tested adding points, badges, and challenges to physical activity. The good part: while the program was running, people walked an average of 1,609 more steps a day. That's about 15 minutes of daily walking that didn't exist before. That's not nothing.
The part almost nobody cites: once the program ended, the effect shrank to just over a third. Something stays. Most of it goes.
The last number in that row is the way out. We'll get there in a moment.
Why it falls off
What you felt in week three has two explanations, and they add up.
The first is novelty. New things are engaging because they're new, and they stop being engaging when they stop being new. Rodrigues and colleagues (2022) followed 756 students for fourteen weeks, in an educational setting rather than a sports one, so take it with a grain of salt. Interest started dropping in week four. But something unexpected happened: between weeks six and ten, it rose again. Those who made it through the slump started using the system differently, no longer out of curiosity but out of habit. The problem is how many people make it through the slump.
The second explanation is more uncomfortable. Deci, Koestner, and Ryan (1999) reviewed 128 experiments and found that expected rewards reduce people's own desire to do the activity. If someone pays you to do something you already enjoyed, once they stop paying, you do it less than before they started.
Put the two together and you get the worst-case scenario for a fitness app: a points system that hooks you for three or four weeks and, along the way, leaves you less eager to run than you were when you installed it.
The way out: it wasn't a point, it was a person
This is where the story changes tone.
There's a theory in psychology, Ryan and Deci's self-determination theory, that explains what sustains motivation once the novelty is gone. It comes down to three things: autonomy (I choose it), competence (I'm improving and I can tell), and relatedness (someone else is with me). Teixeira and colleagues (2012) reviewed 66 exercise studies, and the conclusion was clear: the people who keep exercising long-term are the ones who do it by their own choice, and above all the ones who feel they're getting better.
And then there's the trial that made me change Dromo's design.
In Patel and colleagues' (2019) STEP UP trial, published in JAMA Internal Medicine, 602 adults were randomly split into four groups: one with no intervention and three with games that involved other people, whether to support each other, collaborate, or compete. All three game groups moved more during the 24-week program. That much was expected.
The surprise came later. Twelve weeks after the program ended, with no more points or prizes, the competition group was still walking 569 more steps a day than the group with no intervention.
The effect hadn't gone away. Because what sustained it wasn't a point: it was another person.
How it translated into Dromo
With all of that on the table, the temptation was still there. The easiest way to make an app look good in its first weeks is to hand out points for everything: for opening it, for completing your profile, for logging in three days in a row. The week-one numbers look beautiful. And week six arrives anyway.
So every mechanic had to justify which of the three needs it served, and what risk it brought:
| Mechanic | What it's for | What can go wrong |
|---|---|---|
| Rewards hidden around the city | Autonomy: you choose where to run | The reward becomes the reason |
| Points per kilometer run | Competence: they measure what you did | Points lose their meaning |
| Rarer rewards as you progress | Competence: milestones you notice | Frustration, running just for the reward |
| Weekly league with other runners | Relatedness: someone else is there | Anxiety, toxic comparison |
| Moving up and down divisions | Relatedness: your position matters | Running out of fear of losing |
| Routes that take you to new areas | Autonomy: territory to discover | Nothing obvious, for now |
That led to three decisions.
1. No points without running
In Dromo, points come from kilometers. There are no points for opening the app, completing your profile, or logging in several days in a row. Map rewards add a bonus, but only if you run to them: they're a detour on your route, not a shortcut.
If Deci is right and rewards compete with the desire to run, the reward should at least tell you how much you ran and nothing else. It's the difference between a score that informs you and one that keeps you busy with tasks that aren't running.
2. A city that changes every week
If the problem is that novelty runs out, the answer was to keep it from running out.
The rewards on the map aren't placed by hand, and they don't stay put. Every week, the system looks at where people are running, adds more rewards in areas with more runners, retires the ones that expire, and makes new ones appear in different places. Next week's city isn't this week's city. Novelty stops being something that happens on launch day and becomes something the system produces all the time.
The route recommender follows the same idea: part of its criteria rewards routes through streets you've never set foot on. It exists solely so that Sunday doesn't suggest Tuesday's loop.
It's the biggest piece of engineering in the project. It's also a bet, which is why, further down, I explain how I'll know whether it works.
3. Small, rotating competition
STEP UP says competition was what held up best. But giant leaderboards have a well-known problem: almost everyone ends up in the middle, nobody sees them, and they leave.
That's why Dromo's league has 35 runners and lasts one week. When it closes, the top runners move up a division, the bottom ones move down, and everyone is randomly reshuffled into new leagues. Nobody always competes against the same people, nobody gets stuck behind the same unbeatable runner, and the finish line is never more than seven days away.
What still doesn't sit right with me
If I only told you what went well, this would be advertising.
Dropping a division. It works because losing hurts more than winning feels good. That's exactly the reason to be wary. A mechanic that hooks people through fear of losing is charging in anxiety what it couldn't earn in interest. It stays for now because the league needs consequences, but it's first on the list for review if I see people running out of fear instead of desire.
The streak reminder. If your streak is at risk, Dromo sends you a reminder on Friday afternoon. One. The line between reminding and nagging is thin, and every app crosses it a little more than the last. One reminder a week feels like the right side of it. Three doesn't.
How I'll know if it works
Week one tells you nothing: every app looks good in week one. What I'll be watching is what happens between week four and week ten, exactly where the evidence says everything falls apart. Three concrete questions:
- Do people keep running as much after week four? Is there a slump? And does the recovery Rodrigues found show up, or does the slump just keep going?
- Do points come from running or from hunting rewards? If rewards start to outweigh kilometers, the mechanic has swallowed the activity and needs adjusting.
- Does the changing city make any difference? It's a cheap experiment: freeze the map in one area for two weeks and compare. If nothing changes, that's a lot of engineering for nothing, and I'd rather know early.
And one question no metrics dashboard can answer: are more people running, or are the same people looking at more screens?
The app you'll stop opening
Let's go back to the app sitting forgotten on your phone.
The evidence says two things at once. That games really do get people moving: 1,600 steps a day is not a minor detail. And that points aren't what sustains the habit. What sustains it is running becoming worth it on its own.
So this is the ending I want for Dromo, and I know it sounds strange coming from the person building it: that a year from now, someone stops opening it, just like you stopped opening yours. But that this time, on that day, they go out for a run anyway.
If you have one of those abandoned apps, tell me what week you quit it and what was the last thing that made you open it. And when Dromo crosses its own week six, I'll publish the data here, whatever it shows.
References
- Mazéas, A., Duclos, M., Pereira, B., & Chalabaev, A. (2022). Evaluating the Effectiveness of Gamification on Physical Activity: Systematic Review and Meta-analysis of Randomized Controlled Trials. Journal of Medical Internet Research, 24(1), e26779.
- Althoff, T., White, R. W., & Horvitz, E. (2016). Influence of Pokémon Go on Physical Activity: Study and Implications. Journal of Medical Internet Research, 18(12), e315.
- Howe, K. B. et al. (2016). Gotta catch'em all! Pokémon GO and physical activity among young adults: difference in differences study. BMJ, 355, i6270.
- Khamzina, M. et al. (2020). Impact of Pokémon Go on Physical Activity: A Systematic Review and Meta-Analysis. American Journal of Preventive Medicine, 58(2), 270–282.
- Faccio, M., & McConnell, J. J. (2020). Death by Pokémon GO: The Economic and Human Cost of Using Apps While Driving. Journal of Risk and Insurance, 87(3), 815–849.
- Patel, M. S. et al. (2019). Effectiveness of Behaviorally Designed Gamification Interventions With Social Incentives for Increasing Physical Activity Among Overweight and Obese Adults Across the United States: The STEP UP Randomized Clinical Trial. JAMA Internal Medicine, 179(12), 1624–1632.
- Deci, E. L., Koestner, R., & Ryan, R. M. (1999). A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin, 125(6), 627–668.
- Teixeira, P. J., Carraça, E. V., Markland, D., Silva, M. N., & Ryan, R. M. (2012). Exercise, physical activity, and self-determination theory: a systematic review. International Journal of Behavioral Nutrition and Physical Activity, 9, 78.
- Rodrigues, L. et al. (2022). Gamification suffers from the novelty effect but benefits from the familiarization effect: findings from a longitudinal study. International Journal of Educational Technology in Higher Education, 19, 13.