Most coaches teach tracking by sending four numbers and a link to an app. Two weeks later the client's log shows 1,400 calories on Tuesday, nothing on Saturday, and a protein figure that has never once been hit. The client is not lazy. They were handed the end state of a skill and asked to perform it on day one.
Tracking is an onboarding sequence, not an instruction. It takes about a month to teach properly, and the first week of it has no targets in it at all.
Decide whether this client should track at all
Before the sequence, the screening. Tracking food is a useful teaching tool for most clients and a genuinely bad idea for some, and the moment you teach it is the moment you find out which.
Do not teach tracking to a client with a history of an eating disorder, a client currently in treatment for one, or a client whose intake form or first conversation describes long periods of restriction, purging, or a pattern of eating rules that has cost them relationships or events. That is not a judgement call you are qualified to make and it is not one you have to make — those clients belong with a dietitian or their GP, and you coach the training while someone else handles the food.
Be cautious with two more groups. Adolescents, where measured intake is rarely the right tool and a parent should be in the conversation. And clients who describe food in moral terms in your first call — "I was good all week", "I cheated on Friday" — which is not a disorder but is the soil one grows in. With those clients you can still coach nutrition; you do it with habits, food quality and meal structure rather than a gram target.
There is a version of this conversation that does not sound alarming. "Before we set numbers, has counting food ever been something you've struggled to stop doing?" is a question a client can answer honestly, and it costs nothing to ask.
The four-week ramp
The sequence adds one number a week. Each week the client only has to get good at one new thing, and each week you are checking a specific skill before you build on it.
| Week | Client does | Target | You are checking |
|---|---|---|---|
| 1 | Logs everything, changes nothing | None | Do they log at all, and every day |
| 2 | Keeps logging, adds a protein target | Protein | Can they find protein in real food |
| 3 | Protein plus a calorie target | Protein, calories | Are the entries roughly accurate |
| 4 | Full split | Protein, calories, carbs, fat | Whether the detail is worth it |
Week one: log, do not change anything
Tell the client explicitly that this week is not a diet and there is no number to hit. Their job is to log everything they eat, including the drinks and the mouthfuls, and to change nothing about what they eat. Telling them not to change anything is what makes the data usable — a client who starts eating differently in week one gives you a picture of a week that will never happen again.
Your job is to look at whether they logged, not what they logged. Seven days of imperfect entries is a pass. Four days is a conversation about when in the day they are logging, because the usual fix is logging as they eat rather than reconstructing dinner at 10pm.
At the end of week one you have their actual maintenance intake, which is a far better starting point than a calculator. Turning that into a working number is setting a client's calorie target.
Week two: add protein only
Protein first, for the same reason it is the first number you set: it is the one that changes body composition and satiety most, and it is the one clients are most often far below without knowing. Give the gram target and nothing else. Do not mention calories yet.
This week teaches food knowledge. A client who has to hit 140g of protein learns within days which foods carry it and which look like they should and do not, and that knowledge survives long after they stop tracking. Ranges by goal and body weight are in how to set protein targets for clients.
Week three: add calories
Now the client has two numbers and a real constraint, because protein and calories together force choices. This is also the week the log first gets interesting to audit: a client hitting calories exactly every single day usually has wrong entries rather than excellent control.
Week four: the full split
Add carbohydrate and fat. For plenty of clients this is where you stop and never add them at all — if calories and protein are consistent and the scale is moving, the split is fine-tuning. Add it when training performance matters, when the client is competing, or when they ask. What the numbers actually are is covered in the glossary piece on macros.
Scales or eyeballing
The kitchen scales question decides how accurate the whole exercise is, and it has two defensible answers.
Scales, for the first month at least. Not forever, and not for every meal, but a client who has never weighed food does not know what 30g of oats or 150g of chicken looks like, and their estimate will be wrong in a consistent direction. A month of weighing calibrates the eye, and after that they can estimate most things and weigh the ones that matter.
Eyeballing from the start, for the client where scales are the thing that will make them quit — someone who eats most meals out, someone who cooks for a family, someone for whom the ritual of weighing is the risk you screened for above. Their data will be less precise. That is a fair trade for data that exists.
What you are not choosing between is accuracy and nothing. A client estimating consistently still gives you a trend, because the error is roughly the same size every week.
The errors that make a log wrong
Five mistakes account for most of the gap between what a client logs and what they ate. Teach them in week one and you spend the next three weeks coaching food instead of correcting arithmetic.
Raw versus cooked. 100g of raw chicken is not 100g of cooked chicken; the water has gone. A client who weighs raw and logs a cooked database entry undercounts by a wide margin. Pick one and be consistent — raw is easier for meat, cooked is easier for rice and pasta.
Oils, butter and sauces. The tablespoon of oil the pan gets is often a bigger number than the snack the client agonised about. Cooking fats and dressings are the single most commonly missed entry, and they are missed because they do not feel like food.
Barcode entries that are somebody's guess. Most food databases accept user submissions, so a scan can return a number a stranger typed in wrong. Teach the client to sanity-check against the packet: if the app says a chocolate bar is 90 calories, the app is wrong.
Weekends. Monday to Friday is often tidy and Saturday is missing entirely, which is precisely where the week's calories went. A client who cannot face logging a night out can still log it roughly — an approximate entry beats a blank day, because a blank day makes the whole week unreadable.
Guessing at restaurant meals badly. Restaurant food carries more oil and butter than the same dish at home. The working convention is to log the closest database match and add a bit, rather than log the dish as if you had cooked it.
In Fitsly the client's log sits on their profile alongside their targets, so you can open a week and see which days are blank, which entries look wrong and whether the protein target was ever close — which is the review this whole sequence depends on.
Aim for 80–90%, not perfect
Set the expectation early: hitting targets on most days most weeks is the goal, and no coach worth paying expects seven out of seven. Household scales, food labels and database entries all carry error, so precision beyond a few per cent is invented anyway.
Say the number out loud to the client, because a client without a stated standard invents one, and the one they invent is 100%. That is the standard that produces the client who declines dinner with friends because they cannot log it, and then quits three weeks later.
The practical version: agree a variance band, judge the week rather than the day, and treat one badly-tracked day as a normal week rather than a restart. Building that flexibility in deliberately is flexible dieting, and reading the resulting data across weeks rather than days is assessing nutrition progress.
The client who tracks too well
The client who forgets to log is an accountability problem and you already know how to coach it. The client who tracks obsessively is a different problem and it does not look like a problem, which is what makes it dangerous. Their adherence is perfect. They are, on paper, your best client.
Logging food before eating it so the day can be adjusted, distress when a meal cannot be measured, refusing social events over untrackable food, weighing everything including foods with negligible calories, driving intake below the target you set, or messaging you about a 40g overshoot as though something has gone wrong.
What you do is not "tell them to relax". Say plainly what you have noticed, without diagnosing: "You've been under target four days this week and you messaged me twice about being slightly over. I want to change how we're doing this." Then change it — pull the targets back to calories and protein, or off tracking entirely and onto meal structure, and hold the scale weigh-ins to once a week or stop them.
If it does not settle, refer. You are a fitness coach; disordered eating is outside your scope, and the referral is the professional response rather than a failure. Say that you want them working with someone who specialises in it, offer to keep coaching the training, and mean it.
Plan the exit before you start
Tracking is a teaching tool, not a lifestyle. For most clients it should end. Say so at the start, because a client told this is temporary approaches it completely differently to one who assumes they have signed up to weigh food forever.
The client is ready to come off it when they can estimate a day within a reasonable margin, when their weeks are consistent without them consulting the app, and when they can describe what a day that hits their targets looks like without doing arithmetic.
Wean rather than stop. Drop to tracking protein only, then to tracking three days a week, then to a single check week each month or before you change a target. Keep the app installed. A client who has learnt to track can come back to it for a fortnight when something needs measuring, which is what the skill was for.
The exception is the client for whom the numbers are genuinely load-bearing — a physique competitor mid-prep, or a client whose appetite is suppressed enough that intake needs watching rather than limiting, which is common for clients on GLP-1 medications. Even there, the detail should be the lightest that does the job.
