You look at yourself every day, which is exactly why you can't tell. Six weeks of real training moves your shoulder-to-waist ratio by about 0.05. Your memory of last month is not that precise, and neither is the mirror.
Six weeks. One of them you didn't scan — so it stays empty, rather than being quietly closed up.
The app frames it for you: phone at chest height, held level, the same spot you stood last week, with last week's outline on screen to line up against. Identical beats perfect — the product measures change, so repeatable conditions matter far more than good ones.
Every photograph is read three separate times and the readings are voted. You are shown the spread, not just the answer, so you can see how much to trust a number without reading a word.
The two photographs are compared directly, never measured apart and subtracted — because the error in a single measurement is larger than a week of genuine change. If the light or your stance didn't match, you are told that instead of being given a result.
A tight cluster and a wide scatter look completely different at a glance. That is the point.
It is being built now and tested on a small number of phones. Leave your email and you'll hear once, when there's a place for you — not otherwise.