Almost everyone who quits a food diary quits in the first fortnight, and usually not because the maths was wrong. They quit because logging a meal cost more than the meal was worth: unlock the phone, find the app, wait for it to load, search a database of thirty near-identical entries, pick one, guess a portion, save.
Do that three times a day and the friction wins. Not on day one — on day nine, at a restaurant, when you decide you will add it later and then do not.
Why the chat app matters
The interesting thing about running a tracker inside Telegram is not novelty. It is that the app is already open.
There is no install, no account, no password and no separate icon to remember to tap. The tracker is a conversation, in the same list as everything else you check dozens of times a day, and the meal you were going to log later gets logged now — which is the only version that counts.
That removes several failure points at once:
- No install step. The largest drop-off in any tracker's funnel is the download itself, and it happens before anyone has seen whether the thing works.
- No login. No email, no password, nothing to lose access to on a new phone.
- No cold start. Opening a chat is instant in a way that opening an app is not.
- It travels. Phone, tablet, laptop — the same log, because it is a chat.
What logging actually looks like
Photograph the plate. A vision model reads it, names the dish, and returns an estimate for calories and macros. You correct anything that looks wrong and save. That is the whole loop, and it takes about as long as it takes to describe.
For the times a photo is awkward — a dark restaurant, a plate already half eaten, a hand busy with something else — you can hold the microphone and say what you ate instead. It is transcribed, understood and logged, and you get the estimate back with a delete button if it misread you.
Neither route asks you to search a food database, which is the step people actually hate.
Where estimates stand
An estimate from a photograph is an estimate. It reads what is visible and infers the rest, so it is good at identifying dishes and reasonable at portion size, and it cannot see the oil already absorbed into the rice.
We wrote up how accurate it really is without rounding the answer in our own favour, because a tracker that oversells its precision teaches people to distrust their own numbers, and that is worse than being slightly off.
For most people, most of the time, an estimate logged every day beats a precise figure logged on Tuesdays. Consistency is doing nearly all of the work here, and the design follows from that.
Getting the target right first
None of this replaces knowing what you are aiming at. Before the logging habit is worth much, you need a calorie target that reflects your body and your week: start with BMR and TDEE, then pick a deficit you can hold.
The tracker's job is to make the daily part cost almost nothing. That is the only part that has to survive twelve weeks.