The short answer#
Take what you currently eat in an average week, and eat a few hundred calories a day less than that. That is the whole method, and it is better than any formula, because it starts from a measurement of you rather than an average of people shaped like you.
The catch is that almost nobody knows what they currently eat in an average week. So the usual advice starts somewhere else.
Where the usual number comes from#
Type your height, weight, age and sex into any calculator and it runs an equation — usually Mifflin-St Jeor — that predicts your resting metabolic rate. Then it multiplies that by an activity factor you chose from a dropdown, and calls the result your maintenance: the intake at which your weight would hold steady.
Two things are worth knowing about that number.
It is a population average, and you are one person. The equations are fitted to groups. Individual resting rates scatter around the prediction by roughly plus or minus 10%, and that is before the activity multiplier — a single dropdown standing in for everything you do between waking and sleeping. Two people who answer every question identically can genuinely need several hundred calories a day apart.
Nothing in it is measured, so nothing in it updates. The formula gave its answer before the app had seen you eat anything, and it will give the same answer next month.
None of which makes it useless. It is a reasonable place to start when there is nothing else. It is a poor place to still be three months later.
What your own weeks can tell you instead#
Energy balance is arithmetic. If you know roughly what you ate over a stretch of days, and you know which way your weight actually moved over the same stretch, you can solve for the maintenance figure that would explain both:
maintenance = average intake − (weight trend in kg/day × 7700)
The 7700 is the energy in a kilogram of body mass. The weight trend is the slope of a line through your weigh-ins, not the difference between two mornings.
This is worth more than the formula for a specific reason: it absorbs the things a formula cannot see. Metabolic adaptation to a long diet. A job that has you on your feet. And, unavoidably, the fact that most people under-record what they eat by something like a tenth — because if your diary is short by 10% and your weight is holding, the arithmetic sees a lower maintenance and hands back a lower budget, which is the correct answer to the diary you actually keep.
When not to trust it either#
The method stands entirely on the diary. If the diary is thin, the number it produces is not cautious — it is wrong in a direction that matters, because a maintenance figure that comes out too low produces a budget below what you need.
So the honest version has refusals in it. SmartCrunch will not print a learned maintenance figure unless all of the following hold:
- at least 10 days in the window carry a food diary
- the weigh-ins span at least 14 days end to end
- at least 85% of the days in the window are logged
- the result lands between 0.8 and 1.3 times a conventional estimate
Miss any of them and the app uses a formula-based estimate instead — and says which one it is using. A number with nothing behind it is worse than no number.
There are five such estimates, tried in order, and the whole ladder is written out on the methodology page along with why each threshold is the number it is.
The floor, and why it exists#
Whatever the arithmetic says, a daily budget has a bottom. SmartCrunch sets it at 1200 kcal: if your goal and your pace would together ask for less, the app clamps the budget there and tells you the deficit you asked for was declined.
The reason is not caution for its own sake. Below that point it becomes difficult to get enough protein and enough micronutrients out of the food you have room for, and the weight you lose starts including more of the muscle you would rather keep. If your target requires a budget under the floor, the target needs a longer timeline, not a smaller plate.
Picking a rate#
A deficit of 300–500 kcal a day is a rate most people can hold for months without arranging their life around it. Bigger deficits are not forbidden and they are not magic: they trade adherence for speed, and adherence is the part that decides the outcome.
Whatever you pick, the same arithmetic runs in reverse to give you a date. If the date is unacceptable, the honest levers are the pace and the target, in that order.
What none of this does#
It does not tell you what to eat. A calorie budget is a constraint, not a diet — it has nothing to say about protein timing, food quality, or whether you should be eating bread. Those questions are real; this is not the tool that answers them.
And it is not medical advice. If you are pregnant, managing a condition that touches on eating, or have a history of disordered eating, the number a phone calculates for you is not the number to act on. Ask someone qualified.