Should You Log Food Before You Eat It?
Short answer
Yes — log the parts of the day you already know about before you eat them, and log the rest as it happens. A log written after the fact is a report card: accurate, sometimes useful, and completely unable to change the number it reports. A log written in the morning is a plan, and a plan can still be edited at four in the afternoon when there is still something you can do about it.
That is the whole argument. Everything below is mechanics and caveats.
What actually changes when the log comes first?
The information arrives while it is still actionable.
Log retrospectively and the day's total is a discovery you make at 23:00. You find out you were 400 kcal over after the last opportunity to be 400 kcal under has passed. Repeat that four times a week and you have a deficit on paper that isn't one in the fridge — the pattern described in tracking calories but not losing weight.
Log in advance and the same 400 kcal shows up at breakfast, when the evening meal hasn't been cooked yet. You can serve rice instead of chips, or the same chips with a smaller piece of salmon, or leave the plan alone and accept the day. All three are decisions. Only one of them is available at midnight, and it's the one where you shrug.
There's a second, quieter effect. Typing a meal out before you eat it is a short pause between deciding and doing, and some of what we eat doesn't survive that pause. Not because the number scolds you — because you noticed.
What does sketching a day look like?
Five minutes, usually with coffee, and only the parts you genuinely know.
A worked example. Say the target is 2,100 kcal.
Breakfast, already decided: 50 g porridge oats with 250 ml semi-skimmed milk and a banana. About 180 + 115 + 105, so roughly 400 kcal. Published figures of this kind come from databases like USDA FoodData Central (<https://fdc.nal.usda.gov/>); treat them as good averages, not measurements of your bowl.
Lunch, a thing you have most days: two slices of wholemeal bread, tinned tuna, a spoon of mayonnaise, an apple. Call it 500 kcal.
Afternoon coffee: a flat white, about 120 kcal.
Dinner, roughly planned: chicken, rice, vegetables, oil in the pan. Call it 700 kcal and leave the detail until you cook.
That is about 1,720 kcal planned, which leaves roughly 380 unspoken for.
The unallocated remainder is the most important part of the sketch, and it is not a reserve you are supposed to finish the day without spending. It is the biscuit in the meeting, the second beer, the handful of your kid's chips, the fact that the dinner portion came out bigger than the plan said. Days contain those. A plan with no room for them isn't a plan, it's a trap.
If your evening is usually the loose part of the day, put more of the slack there. If your mornings are improvised and your dinners are fixed, do the opposite.
How do you adjust against what's left?
At some point in the late afternoon, look at the remainder rather than the total. That is a different and much easier question: not "how did I do" but "what fits".
Say 1,000 kcal is left and dinner as planned is 700. Nothing to do. Say 450 is left because lunch was a pub lunch. Then dinner is the same chicken and vegetables with less rice and less oil, which is the lever most people underestimate — a generous glug of olive oil is around 120 kcal a tablespoon, and sauces and cooking fat are where portions quietly inflate.
And sometimes the honest adjustment is none: you go over, you log it, the week absorbs it. A single day being 300 over is noise against a week. The weekly arithmetic is what moves the scale, and a kilogram of body fat is about 7,700 kcal — nothing you do on a Tuesday evening is that.
Where pre-logging goes wrong
Three failure modes, all common.
The plan is simply wrong. You planned 700 for dinner and the restaurant served 1,400. Pre-logging a meal is a forecast, and forecasts made in the morning about a plate you haven't seen are frequently out by a few hundred calories. If you pre-log, you have to go back and correct the entry once the food is real. A plan you never reconcile is fiction with a timestamp.
Rigid plans break on contact with real days. Someone brings cake. The train is delayed and dinner is a petrol station. If your plan has no slack, the first deviation makes the whole log feel void, and "the log is void" is one step from "I'll start again Monday". Getting back on is doable — see starting again after a break — but it is better not to engineer the fall.
For some people the plan becomes a rule they've failed. This is the serious one. If writing the day down in advance turns eating into a compliance test, and eating off-plan feels like a moral event rather than a Tuesday, pre-logging is making your life worse and you should log afterwards instead. The method is supposed to lower the stakes of a decision, not raise them.
One thing worth saying plainly: pre-logging is not a licence to eat nothing all day so one meal fits. Deliberately front-loading restriction to bank calories for the evening tends to produce an evening you planned very carefully and then didn't recognise. Spread the day out; let the remainder be a cushion, not a hoard.
The hybrid almost everyone lands on
Not a fully planned day. Not a fully retrospective one. The anchored parts pre-entered, the variable parts logged live.
Most people's food is more repetitive than they think. The same breakfast four days in five, the same lunch at a desk, the same batch of chilli from Sunday. Those get entered in the morning in under a minute, often by copying the same meal from a previous day — and if it came out of a big pot, the per-portion figure only has to be worked out once. The dinner you haven't cooked, the drinks you might have, the thing a colleague brings in: those get a rough placeholder or nothing at all, and the real entry when it happens.
This is the shape Day So Far is built around, because describing a planned meal in a sentence — "chicken thigh, rice, broccoli, a bit of oil" — takes about as long whether the food is in front of you or still in the freezer, and the estimate comes back the same either way. It's an estimate in both directions, and a planned meal is an estimate of an estimate. Plan with it; correct it afterwards.
The test of whether this is working isn't whether your plan matched your day. It's whether you knew where you stood while there was still time to do something with the information.