You set a 500-calorie deficit. You have been careful for three weeks. The scale has not moved.
The internet will offer you starvation mode, a broken metabolism, cortisol, or a thyroid problem. Occasionally one of those is the answer. Far more often the deficit is simply smaller than the log says, and the reasons are boringly specific.
Here they are, roughly in order of how often they turn out to be the culprit.
1. Cooking oil
This is the big one, and in Indian and South Asian kitchens it is the big one by a wide margin.
A tablespoon of oil or ghee is about 120 calories. A family dal gets two or three tablespoons in the tadka. A sabzi gets two. A paratha absorbs one or two on the tawa. None of it is visible on the plate by the time you sit down, and none of it appears if you log "dal, sabzi, two rotis" against generic database entries built for a different kitchen.
Three unlogged tablespoons a day is 360 calories. That alone converts a 500-calorie deficit into a 140-calorie one, which produces roughly 0.6 kg a month — indistinguishable from noise on a bathroom scale.
What to do: for one week, measure the oil that goes into the pan and divide by the number of servings. You do not need to keep doing it. You need to find out once whether your estimate was 1 teaspoon or 3.
2. Liquid calories
Chai at 60–90 calories a cup, four times a day, is 240–360 calories that nobody thinks of as eating. Add a sweetened coffee, a glass of juice or a lassi and you are past 500 before any food is involved.
The chai page has the per-cup ranges. The point is not that chai is bad — it is that four cups is a meal's worth of energy arriving in a form that does not register as one.
What to do: log drinks for a week. Just the drinks. Most people are surprised once and then adjust permanently.
3. Nuts, and anything eaten by the hand
Almonds are 570–610 calories per 100 g. A "handful" is anywhere between 15 and 35 of them. Peanut butter is 600 per 100 g and the heaped spoon most people use is nearly double a level tablespoon.
Calorie-dense food eaten without a serving boundary is the single most reliably under-logged category there is, because there is no natural stopping point and no unit to count.
What to do: count them out into a bowl rather than eating from the packet. This is not about discipline; it is about creating a serving where none exists.
4. Weekends
Five careful days at a 500-calorie deficit is 2,500. Two relaxed days at 800 over maintenance is 1,600. Net weekly deficit: 900 calories, or roughly 0.1 kg.
The weekdays were fine. The week was not. This is the most common pattern in people who are convinced their diet is working and their body is not.
What to do: log the weekend, even roughly, even badly. You do not have to change it — you have to be able to see it. Most people correct on their own once the week is visible as a week.
5. Restaurant food
A restaurant kitchen's job is to make food taste good, and fat is how that is done. The same dish that is 400 calories at home is routinely 700 out. Portions are larger, the oil is more generous, and there is often butter or cream at the finish that never appears in the description.
Twice a week, that gap is around 600 calories. Tracking calories when eating out covers how to estimate it without ruining the meal.
6. Your TDEE estimate was optimistic
Activity multipliers are the weakest part of any calorie calculator. Choosing "moderately active" when you train three times a week but sit for the rest of the day can overstate your burn by 300–400 calories, which is most of a deficit before you have eaten anything.
What to do: if the log looks genuinely accurate, this is the next suspect. Drop the target by 150–200 calories and give it another three weeks. See TDEE vs BMR for why the multiplier is so unreliable.
7. You are actually losing fat and the scale is hiding it
Body weight moves for reasons unrelated to fat: sodium, carbohydrate intake, water retention from a hard session, hormonal cycles, and simply what is currently in your digestive tract. Two kilograms of day-to-day variation is entirely normal.
This means a real 0.4 kg-per-week fat loss can be completely invisible for two or three weeks and then appear all at once.
What to do: weigh daily under identical conditions — morning, after the toilet, before eating — and read the weekly average rather than any single number. Judge nothing on fewer than three weeks of data.
What is almost never the answer
"Starvation mode." Metabolic adaptation is real and modest — roughly 5–10% beyond what is explained by being smaller. It does not stop weight loss and it does not make a genuine deficit produce zero results.
Eating too little. At a normal deficit this is not a mechanism for failing to lose weight. Very low intakes cause problems, but they are muscle loss, misery and poor adherence — not weight stalling.
A specific food. Rice does not prevent fat loss. Nor does fruit, nor eating after 8pm. The deficit is the mechanism; the composition changes how pleasant it is, not whether it works.
The order to check things in
- Log everything for one honest week — oil, drinks, tastes while cooking, weekend included.
- Weigh daily, compare weekly averages, ignore individual days.
- If the week's average intake really was 500 below your estimated TDEE and the trend is flat over three weeks, lower the target by 150.
- If it stays flat with genuinely accurate logging over six to eight weeks, that is the point at which a doctor is a reasonable next step rather than a first resort.
Nine times out of ten it stops at step one.
The reason step one is hard is that honest logging is tedious, and tedium is what makes people stop. BetterCal is built around removing exactly that: describe the meal in a sentence, correct anything that looks wrong once, and repeats are a single tap. The point is not perfect numbers — it is having a week of real data to look at when the scale disagrees with you.
Related: what is a calorie deficit · why people quit calorie tracking · calorie deficit calculator
