Most coaches add habits to a client's plan the same way they add accessory work: a list of good things, all switched on at once, on day one. Drink three litres. Ten thousand steps. Bed by ten. Two weeks later the client has stopped logging all of them and neither of you mentions it again.
The failure is almost never the client's discipline. It is that the habits were chosen by the coach, there were too many of them, and the first one was too big to complete on a bad day.
This covers how to choose the habit that actually matters, how to size it, how many to run at once, how to track it without becoming the person who checks up, and what to do at the two moments that decide whether habit coaching works at all — the broken streak and the habit that has stopped being hard.
The habit has to be the client's choice
Start here, because everything downstream depends on it. If you pick the habit, you own it, and the client is complying with your homework rather than changing anything about how they live. Compliance survives about three weeks of enthusiasm and then quietly stops.
That does not mean you have no opinion. It means the order of the conversation changes. You ask what gets in the way — not "what habits do you want to build", which produces the answer the client thinks you want to hear. Ask what a bad week looks like, what time they eat, when they sleep, what happened the last three times they missed a session. Then you propose two or three candidates from what they told you and let them pick one.
The client choosing badly is fine. A client who picks the wrong habit and does it for a month has learned they can do a thing they said they would do, which is most of the point. A client who follows your correct habit for nine days and stops has learned the opposite.
Choose the habit that unblocks the others
Given a free choice, most clients pick the most virtuous-sounding option. Meditation, cold showers, a 5am alarm. These are rarely the constraint.
Look for the habit that, if it happened, would make two or three other things happen by themselves. It is usually unglamorous and usually earlier in the day than the problem it fixes.
A client who cannot get to the gym after work has a leaving-work problem, not a motivation problem. A client whose evening eating is out of control has a lunch problem. A client who misses the morning session has a bedtime problem. The habit worth running is the one upstream of where the symptom shows up.
Two questions find it quickly. First: when this goes wrong, what happened before it went wrong? Second: is there something that, if it were already done, would make this decision unnecessary? Packing the gym bag the night before is not a wellness habit, it is the removal of a decision at 6am from a person who is bad at decisions at 6am.
Be honest that this sometimes means the first habit has nothing obviously to do with training. A client who is not sleeping does not need a step target.
Make week one small enough that the client cannot fail
Whatever number the client proposes, they will propose it on a good day, imagining a good week. Your job is to size the habit for their worst realistic week — the one with the sick kid and the work deadline.
Take the client's proposal and cut it until they laugh at how easy it is. Ten minutes of walking, not 10,000 steps. One glass of water with each meal, not three litres. Two vegetables at dinner, not a tracked macro target. The point of week one is not the physiological effect, which at that size is nil. The point is a full week of completed days, because the thing you are actually building is the client's belief that when they say they will do something, they do it.
The size test: could they do this on the worst day of the month? If not, it is too big. A habit that only works on good days is a mood, not a habit.
Then leave it there for longer than feels useful. The urge to raise the target the moment a client strings together five green days is the same urge that adds 10kg to a set that moved well, and it has the same failure mode. Two clean weeks before you touch the number.
How many habits at once? Usually one
The honest answer is one. Two if they are genuinely the same behaviour at different times of day.
Coaches resist this because one habit looks like a thin service, especially at a premium price. Say out loud why you are only giving them one: three habits at once means three chances to fail per day, and a client who fails one usually abandons all three. There is a second reason — if a client is running four habits and progress stalls, you cannot tell which one did nothing.
The exception is the client who is already consistent and wants more, and even then you add the second habit after the first has run for a month without you mentioning it. If you still have to remind them, it is not established, and adding to it is adding to sand.
Common problems and the habit that addresses them
The mapping below is not a prescription — the client still picks — but it gives you the two or three candidates to put in front of them, and it points upstream of the symptom rather than at it.
| What the client reports | The habit that usually addresses it |
|---|---|
| "I have no energy for evening sessions" | A fixed lunch, eaten away from the desk |
| "I eat well all day and then lose it at night" | Protein at breakfast and lunch, no target for dinner |
| "I keep missing morning workouts" | Lights out at a set time, four nights a week |
| "I'm too busy to train" | Session booked in the calendar like a meeting, night before |
| "I do the sessions but nothing's changing" | Log every session in full, including the ones that went badly |
| "I snack constantly while working" | One planned afternoon snack, eaten away from the screen |
| "My weekends undo my week" | Saturday and Sunday breakfast, same as a weekday |
| "I can't get 10,000 steps" | A ten-minute walk after one specific meal |
| "I'm always sore and never recovered" | A set wake time, including weekends |
| "I forget to track anything" | Log one thing, at the same time, daily |
The pattern in the right column: each one is tied to something that already happens at a fixed time. Habits attached to an existing anchor — a meal, the commute, putting the kids down — survive far better than habits attached to a feeling or a spare moment, because "when I get a chance" is not a time and never arrives.
Track it without turning into surveillance
Tracking is what separates a habit you are coaching from a suggestion you made. It is also the part that goes wrong fastest, because the same daily log that gives you information gives you the opportunity to comment on it, and a client who feels watched every day starts logging for you rather than for themselves.
Three rules keep it useful:
The client logs, you review. Look at the week, not the day. Daily reactions turn you into a supervisor and teach the client that a missed log will produce a message, which mostly teaches them to log dishonestly.
Say nothing about a single miss. One missed day is noise. Three in a week is a signal, and the signal is that the habit is wrong or too big, not that the client needs a reminder.
Never make the log a report card. If the only time habits come up is when they were missed, the client learns that logging invites criticism. Bring up the streak when it is intact, too.
Where the data is already being collected, use it. Steps, sleep and resting heart rate that arrive from a watch cost the client nothing to record and remove the "I forgot to log it" failure — coaching clients with wearable data covers what that data is and is not good for. For everything else, one tap a day is the ceiling. If logging takes longer than doing the habit, the log is the thing they will drop.
The mechanics matter less than the review. What you want in front of you before a check-in is a week per client — which days were met, which were part-done, where the streak sits — rather than a message thread you scroll to reconstruct it. In Fitsly, habits sit alongside training and nutrition on the client's profile: you assign one from a reusable habit library, the client logs it in their app, and their week comes up as a grid you can read before the call instead of during it. Any tool that gets you the same week in one view will do, though a spreadsheet the client maintains is worse than either.
What to do when the streak breaks
This is the moment the whole thing turns on. A client with 23 consecutive days misses a day, and the number resets to zero. What they feel is not "I missed one day" — it is that the 23 days were wasted. That is when clients quit, and it is why streak counters are a double-edged tool.
Get ahead of it. Tell the client at the start, before there is a streak to break, that they will miss days and that missing a day is part of the design, not a failure of it. The rule is: never miss twice. One missed day is a day. Two in a row is the start of not doing it any more.
When it happens, the conversation is diagnostic, not motivational:
- What was different about that day? Usually something structural — travel, a shift change, a sick kid. If the habit cannot survive a normal disruption, the habit is wrong.
- Would a smaller version have survived it? This is the most useful question you can ask. If the answer is yes, you now have the habit's floor: the version they do on bad days, which keeps the chain alive.
- Do we still want this one? Sometimes the honest answer is no. Dropping a habit deliberately is completely different from letting it decay, because the client is still the one deciding.
Build the floor in from the start and the reset rarely happens. A step target with a five-minute walk as its minimum is a habit with a bad-day version. A three-litre water target has no floor, which is why it dies the first time someone is on a plane.
If a client goes quiet on habits for a fortnight, treat it as the early warning it is rather than a habits problem. Silence around the small stuff usually shows up before silence around the training, and why online coaching clients quit starts with exactly this pattern.
When a habit should graduate
A habit is finished when the client stops noticing they do it. They do not think about the lunchtime walk, they just walk. At that point tracking it costs them a daily tap and tells you nothing you did not already know.
Retire it. Say explicitly that it has graduated, that it is now just something they do, and that you are taking it off the list because it is done. Clients hear this as a promotion, which it is, and it clears the space for the next one. Leaving a mastered habit on the board is how a client ends up with eleven daily taps, most of them meaningless, and eventually stops doing any of them.
Three signs a habit is ready to go:
- They have done it for a month without you raising it once.
- They do it on bad days, at the reduced version, without being told to.
- The log has become a formality — they tick it after the fact, from memory.
Then pick the next one the same way you picked the first: ask what is getting in the way now. The answer will be different, because fixing the first constraint has moved it. This is the same logic as progressive overload for online clients — small, deliberate increases to something the client is already handling, rather than a jump to what you think they should be capable of.
What "21 days to form a habit" is worth
Not much. The figure is widely traced to a 1960 self-help book, Psycho-Cybernetics, by the plastic surgeon Maxwell Maltz, who reported that his patients took around three weeks to get used to their changed appearance. It was an observation about surgical patients, not a study of behaviour, and it became a rule of thumb through repetition rather than evidence.
There is real research on how long behaviours take to become automatic, and its main finding is that the range between people is enormous — some behaviours settle in a few weeks and some take the better part of a year. That is more useful to you than any average, because it means a client who is still finding it hard at week six is not behind.
So do not put a number on it with clients. "Give it three weeks and it will be automatic" sets up a deadline they will fail, and the failure feels like evidence about them. Say instead that it gets easier and that you will keep it small until it does.
The mistakes that cost the most
Assigning habits at onboarding. Week one is already a lot of new information. Habits assigned before you know anything about the client's week are guesses wearing a plan's clothes. Wait until after the first or second check-in.
Habits with no time attached. "Drink more water" is an intention. "A glass of water with each meal" is a habit, because it is attached to something that already happens.
Tracking things the client cannot control. Weight is an outcome, not a habit. So is sleep duration, mostly. Bedtime is a habit; hours slept is a result of one.
Adding a habit to fix a habit. When a habit is failing, the answer is a smaller version of that habit, not a second one propping it up.
Only discussing habits when they fail. If the only mention is a chase, you have made the log a surveillance tool. The weekly check-in is the place for this — see how to run client check-ins for the structure that keeps it a conversation rather than an audit.
Confusing habits with accountability. Habits are what the client does when you are not there. Keeping online clients accountable is a separate job, and doing it well is what buys you the room to keep the habit list short.
