A new client fills in the intake form, you put their height, weight, age and activity level into a calculator, and it gives you a number. You subtract 500, round it, and send it over. Three weeks later the scale has not moved and you have no idea whether the number was wrong, the logging is wrong, or three weeks is simply not long enough to tell.
That is the actual job. Picking the first number takes four minutes and is mostly arithmetic. What you do with weeks two, three and four is the part clients are paying for, and it is the part that never appears in a calculator.
The first number is a hypothesis
Treat the starting target as a testable guess about the client's maintenance intake, because that is all it is. Every input has error in it: the equation predicting their resting metabolism is a population average applied to one person, the activity multiplier is a bucket they chose themselves, and the weight they typed in was from whenever they last stood on a scale.
Stack those together and a starting estimate can sit several hundred calories away from the truth in either direction. That is not a reason to skip the calculation — you have to start somewhere — but it does change what you say when you hand the number over. "This is our starting point and we will adjust it in a fortnight" sets up the next four weeks correctly. "This is your calorie target" sets up a client who thinks the plan has failed the moment the first week is flat.
It also changes what you do next. If the first number is a hypothesis, you need a way to test it, which means intake and body weight both have to be recorded from day one. A client who diets for three weeks without logging either has given you nothing to work with, and you will end up guessing a second time.
The full method for arriving at that first estimate — the equations, the activity multipliers, and how to convert the result into grams — is in how to calculate TDEE and macros for clients. The rest of this article assumes you have a number and now have to live with it.
Set a rate of loss, not a calorie figure
Do not start by choosing a deficit. Start by choosing how fast this particular client should lose, express it as a percentage of their body weight per week, and let that decide the calorie number.
A fixed rate in kilograms punishes small clients and wastes time for large ones. Half a kilo a week is a mild deficit for a 120kg client and an aggressive one for a 58kg client, but "0.5% of body weight per week" means the same thing to both.
| Rate per week | What it suits | What it costs |
|---|---|---|
| 0.25–0.5% | Lean clients, long timelines, anyone lifting for strength, clients with a history of yo-yo dieting | Slow enough that a bad week hides the trend; needs patience from both of you |
| 0.5–0.75% | The default for most general-population fat loss | Manageable hunger, training holds up, adherence is realistic |
| 0.75–1% | Clients with a lot to lose, or a genuine short deadline | Hunger and fatigue become real; needs closer contact |
| Above 1% | Rarely justified outside of clinical supervision | Adherence collapses, training quality drops, muscle loss rises |
For most clients, 0.5–0.75% per week is the band worth defaulting to. It is fast enough that the client can see it working within a month and slow enough that they can keep doing it after the novelty wears off.
Why aggressive deficits fail
The usual explanation for a crash diet failing is metabolic — the body "goes into starvation mode" and stops burning fat. That is not what happens, and it is not the reason to avoid a large deficit.
Energy balance still works at 1,200 calories. The client will lose weight. The problem is that they will not keep eating 1,200 calories, because the deficit makes them hungry, tired, socially restricted and irritable, and those four things together break adherence long before physiology has anything to say about it.
Look at what a large deficit actually asks a client to do. They have to say no to food at every meal, decline social events built around eating, train while under-fuelled, and do it for months while sleeping worse. Very few people can hold that. What normally happens is four excellent days, a weekend that undoes them, guilt, and a restart on Monday — a cycle that averages out to roughly maintenance while feeling like constant dieting.
A moderate deficit works because the client barely notices it. Two hundred to four hundred calories below maintenance is a smaller dinner and no second serving. It survives a birthday, a work trip and a bad night's sleep, and it is still there in month four when the aggressive plan has been abandoned twice.
There is a secondary cost worth naming. In a large deficit, spontaneous movement falls off — fewer steps, less standing, less fidgeting — so the actual deficit ends up smaller than the one you prescribed, on top of a client who now feels terrible. You get a fraction of the extra speed you cut for.
The floors that constrain how low you can go
The calorie number is not free to fall as far as the arithmetic wants. Three things put a floor under it.
Protein. Protein has to stay high in a deficit, because it is what protects lean mass and it is the macro that does most to keep a client full. Coaching convention lands somewhere around 1.6–2.2g per kilogram of body weight, higher at the leaner end and for clients training hard. That is a fixed cost in the budget before you allocate anything else — how to set protein targets for clients covers where to sit within that range.
Fat. Fat cannot go to zero. It carries fat-soluble vitamins, it is involved in hormone production, and diets that cut it to nothing are miserable to eat. A common working floor is around 0.6–0.8g per kilogram, or roughly 20% of total calories, whichever is higher.
The total. Once protein and fat are paid for, whatever is left is carbohydrate — and if that number is very small, the target is too low for this client. As a general convention, coaches do not programme below roughly 1,200 calories a day for women or 1,500 for men, and even those are low enough to warrant a conversation rather than an assumption. Below that it becomes genuinely hard to get adequate protein, fat and micronutrients from food, training suffers, and you are in territory that belongs to a clinician rather than a coach.
If your arithmetic demands a target under the floor, the answer is not a lower target. It is a slower rate of loss, more activity, or a longer timeline — usually the first.
When to refer out instead of setting a number
Some clients should not get a calorie target from you at all, and saying so early is easier than unwinding it later.
Refer to an accredited dietitian, or back to their doctor, when a client has a history of disordered eating or an eating disorder, a diagnosed condition where intake matters clinically — type 1 or type 2 diabetes, kidney or liver disease, coeliac disease — is pregnant or breastfeeding, or is taking medication that affects appetite, weight or blood glucose. Under-18s belong in the same category.
This is not a legal disclaimer to bury at the bottom of a page. General nutrition guidance for healthy adults is inside a fitness coach's scope; individualised dietary treatment for a medical condition is not, in most jurisdictions, and a coach who sets an aggressive target for a client with a restrictive-eating history can do real harm. Referring out is a normal part of the job, and clients read it as competence rather than as a coach passing them off. You can still coach their training and their habits while someone else owns the intake.
A worked example
Priya, 34, 82kg, desk job, trains four times a week, wants to lose fat for a wedding in seven months. Estimated maintenance from the calculator: 2,250 calories.
Rate. She has time and lifts seriously, so 0.6% of body weight per week — about 0.5kg. That implies a deficit of roughly 380 calories a day, which rounds to a target of 1,850 calories.
Floors. Protein at 1.8g/kg is 148g, call it 150g (600 calories). Fat at 0.8g/kg is 66g, call it 65g (585 calories). That leaves 665 calories for carbohydrate, about 165g — comfortable, so the target clears the floors.
Now the part that matters. She logs food and weighs herself most mornings, and you look at the weekly averages rather than any single day.
| Week | Avg logged intake | Avg body weight | Days logged | What you do |
|---|---|---|---|---|
| 1 | 1,910 | 81.6kg | 7 | Nothing. Week one is water, not fat. |
| 2 | 1,880 | 81.4kg | 6 | Nothing yet. Two weeks is the minimum readable trend. |
| 3 | 2,050 | 81.3kg | 4 | Logging slipped. Fix the logging before the target. |
| 4 | 1,890 | 81.1kg | 7 | Trend is about 0.25kg/week — half the plan. Cut 150 calories. |
Two things in that table are the whole article. The first is that you did not touch the number for three weeks. The second is that in week three the problem was the data, not the diet — dropping her calories on the strength of four logged days would have been guessing dressed up as coaching. When and how much to move the number once you have a readable trend is covered in how to adjust client macros.
Assume under-reporting, and treat it as a skill gap
Clients under-report what they eat. This is well established, it happens to experienced trackers as well as beginners, and it is almost never dishonesty. Cooking oil goes in the pan and not in the app. The milk in three coffees is invisible. Weekends get logged from memory on Monday. Portions get estimated by eye and eyes are optimistic.
The practical consequence is that when a client logs 1,850 and does not lose weight, you have two candidate explanations: their maintenance is lower than you estimated, or their actual intake is higher than the log says. They call for opposite responses, and cutting the target when the real problem is measurement makes the client hungrier without making them lighter — then it happens again in a fortnight.
Sort them apart before you change anything. Ask how they log oils and drinks, whether they weigh food or estimate, and what happens on Saturday. Have them weigh everything for four days including a weekend day, and compare against the week they logged by eye. If the gap is 300 calories, you have found your answer without touching the target.
Then coach the skill instead of tightening the screw: weigh food raw, log before eating rather than after, save frequently eaten meals so a repeat takes seconds. Teaching clients to track macros is a coaching task in its own right, and it pays off for the whole time you work together. For clients who will not log consistently no matter what you do, a structured plan they follow rather than a number they measure against is often the better fit — how to build a client meal plan covers that route.
Whatever tool you use needs to show you consistency alongside the average. In Fitsly, a client's food log week shows their average daily calories against their goal and how many of the seven days they logged, which is the number you have to read first — an average across four days is not an average of their week.
Tell the client what the scale is going to do
Most of the trouble in the first month is not the target. It is a client who expected a line and got a scribble.
Body weight moves for reasons that have nothing to do with fat. Sodium and carbohydrate intake shift water retention by a kilo or more within a day. Undigested food adds weight until it does not. A hard training session leaves inflammation and water in the muscle for a couple of days. Many women see a predictable rise across part of the menstrual cycle. Poor sleep, travel, alcohol and stress all show on the scale.
So say this at the start, before the first weigh-in rather than after the first bad one:
- Weigh daily, same time, after the toilet, before eating, and treat the weekly average as the real number.
- Expect individual days to go up. A day that goes up is not a failed day.
- A flat week is normal. Two flat weeks with good logging is a signal.
- The first week often shows a fast drop that is mostly water, and the second week often shows nothing. Neither is the plan working or failing.
Half of adherence in month one is the client not panicking, and a client who was warned about fluctuation stops treating each morning as a verdict.
