How to
How to run a fitness challenge your friends will actually finish
Most group fitness challenges die by day nine. The fixes are boring and specific: the right metric, the right length, the right stakes, a fair scoreboard.
Quick answer
Keep it small, short and fair: three to eight people, one week at a time with a Monday reset, scored on a unit every device counts the same way, which means active minutes or kilometres for a mixed group, never steps across different brands. Add a small silly stake framed as something to lose, keep the leaderboard where everyone already looks, and restart every week so a bad Tuesday never ends anyone's season. Formats that reset beat formats that accumulate.
Every group challenge is born the same way: a burst of group-chat enthusiasm, a 30-day plan, eleven participants, a spreadsheet someone swears they will update. And most die the same way, around day nine, when the person in last place stops opening the app and everyone politely never mentions it again.
None of that is a character problem. It is a design problem, and the design is fixable. Here is the recipe, with the reasoning behind each rule.
Rule 1: one metric, with real units
The fastest way to kill a challenge is a scoreboard someone can reasonably call rigged. Steps read differently on a wrist, a ring and a phone, which is why cross-brand step contests wobble, and branded scores like Recovery or Body Battery cannot rank two people at all. For a mixed-device group that leaves a short honest list: active minutes first, since every major brand counts effort against the same public 150-minutes-a-week guideline, then kilometres, then workout time.
Challenge formats, and when each one is fair
| Format | Fair when | Falls apart when |
|---|---|---|
| Active-minutes week | Any mix of brands and fitness levels | Someone has no wearable at all |
| Distance chase (km) | Everyone runs, rides or walks outdoors | Gym-only members feel invisible |
| Step race | Everyone wears the same type of device | Wrists race pockets and rings |
| Ring-close percent | All-Apple groups with honest goals | Anyone owns a different brand |
If you want a format to start from, the fitness challenge ideas list has 40 of them, sorted by the group they suit, and every one is built to fit these rules.
Rule 2: a week, on repeat
The STEP UP trial is the cautionary tale: every challenge design worked while the game ran, and almost everything faded once it stopped. One-off challenges end; that is their flaw. A repeating week fixes both failure modes at once. Nobody is ever more than six days from a clean slate, so a disastrous Tuesday cannot end a season, and there is no finish line after which everyone collectively exhales and stops.
Apple reached the same conclusion from its own data: its native competitions are seven-day rounds, scored as a point per percentage of ring closed, capped at 600 a day. If your group happens to be all-Apple, that format is one tap away and genuinely well designed; here is how the Apple Watch competition works.
Rule 3: three to eight people, matched loosely
The research on social exercise keeps returning one finding: rank motivates when the gap is chaseable. The experiment that roughly doubled how long people exercised paired them with a partner slightly better, not dramatically better. In a group of twenty, the bottom five are out of every race by Wednesday and behave accordingly. Three to eight keeps everyone within shouting distance of someone, which is the actual fuel. The same rule scales up: a work fitness challenge splits the office into small teams, and a virtual fitness challenge splits a scattered group into brackets.
If your group spans a marathoner and a beginner, do not pretend the gap away. Score on effort relative to each body, which active minutes partly do on their own since they key off personal heart-rate zones, or run a ring-percent format where closing your own goals is the game.
Rule 4: stakes that sting a little, pleasures that bundle
Two behavioural tricks carry more weight than any prize. First, frame the stake as a loss: the Penn wearable trials gave players points to defend rather than earn, because losing ten points hurts more than winning ten feels good. In friend terms, everyone puts five dollars in on Monday and the winner takes the pot, or last place buys the coffees. Small enough to laugh at, real enough to feel.
Second, temptation bundling: chain something you crave to the activity you avoid, the trashy podcast you only allow yourself on a walk. The behavioural scientist Katy Milkman tested exactly this with audiobooks held hostage at the gym:
Rule 5: the scoreboard lives where the group already looks
A challenge tracked in a spreadsheet is a challenge one volunteer sustains alone, and it outlives their patience by zero days. The leaderboard has to update itself from the devices and surface where the group already talks, with the day’s rank one glance away. This is also what keeps the contagion loop from the research alive: your crew’s activity has to be visible to pull you, and yours visible to pull them.
That job description is, not coincidentally, Fit, the cross-brand fitness competition app: standing crews of friends, ranked on active minutes and distance across Apple, Garmin, Fitbit, WHOOP and the rest, with the week as the unit of play. It is our app, so the honest caveat comes with it: an all-Apple group is well served by Apple’s built-in competitions, and an all-Fitbit group that just wants steps has simpler options. Fit is for the group you actually have, the one with four brands in it.
What to do with this
- Post one message today: “Weekly active-minutes challenge, resets Mondays, five bucks in, who’s in?” Cap it at eight yeses.
- Set the scoreboard up once, automated, in whatever app fits your group’s devices; the full comparison is the shopping list, and the step challenge apps guide covers step races.
- Protect the bottom of the table. Check on last place by Wednesday; a chaseable gap is your job as organiser, and it is worth more than the pot.
- Restart every Monday without a vote. The challenge that survives is the one that was never allowed to end.
Questions people actually ask
How long should a fitness challenge with friends last?
One week, repeating, with a reset every Monday. Month-long challenges die in the second week because one bad stretch makes the remaining days feel pointless, and the trial data shows effects fading when a one-off game ends. A weekly rhythm means nobody is ever more than six days from a fresh start, which is also why Apple built its native competitions as seven-day rounds.
How many people should be in a step or fitness challenge?
Three to eight. Below three, one dropout kills it. Much above eight, the bottom of the leaderboard stops feeling chaseable and the people who need the challenge most quietly disengage. The motivation research is consistent that rank pulls hardest when the person ahead of you is within reach.
What is a fair fitness challenge metric for different devices?
Active minutes first, kilometres second, workout time third. All are real units that Apple, Fitbit, Garmin and Samsung report on roughly the same scale. Steps only stay fair when everyone wears the same type of device, and branded recovery or readiness scores are never comparable between people at all.
Should a fitness challenge have money or stakes on it?
Small and silly beats large and serious. The behavioural trials that worked framed points as something you start with and lose, not something you slowly earn, because losses sting more than gains please. A five-dollar pot, loser buys coffee, or a dare works; high stakes push people to quit entirely or to cheat the tracker.
Why do group fitness challenges always fall apart?
Almost always one of four killers: the metric was unfair across devices, the round was too long to recover from a bad week, the group was too big to feel close, or the leaderboard lived somewhere nobody looked. Each has a boring fix, and the fixes matter more than enthusiasm, which every group has in week one.
Sources
Get the next one by email
Occasional notes on making fitness numbers comparable across devices, and reading your own baseline. No spam, unsubscribe in one tap.
We email you once to confirm. Your address is only used for these notes.