The smallest food database on the market. On purpose.
Why we don't import millions of products – just the ones people actually eat, verified and current.
Published on
"We have millions of products" isn't a promise, it's a warning.
In most tracking apps, a large product database sounds like a selling point; to us, it's the first sign that something's off. What good is a database full of dead entries no one uses – products that maybe aren't even on shelves anymore?
What a huge database means in daily use
You scan a barcode and get twelve hits: three with different calorie counts, one with a completely different ingredient list, one marked as "verified" – but verified by whom, when, against which source? Doesn't say. So you end up comparing values against the label in your hand, deciding for yourself which entry is probably right. That's exactly the work the app was supposed to take off your hands.
It fails the other way around just as easily: a supermarket chain rebrands, the package contains the same thing as last week, only the design is new – and with the new design, the product picked up a new barcode. You scan, no match. The app that advertises "millions of products" doesn't know the one you're actually holding. Because mass isn't the same as being up to date.
Few products, each with a traceable origin
We do it the other way around: to start with, only base ingredients are in there – oats, carrots, lentils, tofu, olive oil. Just the few hundred building blocks most products are made of in the first place, with values from traceable sources (where we get them is written up here).
Branded products only land in the database once someone submits them and we've verified them – not before. Meaning: what shows up in search has actually been verified, not just labeled as "verified."
Submitted by people who actually eat the product
The decisive difference compared to imported bulk data: every product in our database is in there because someone submitted it, and nobody submits a product they don't put in their own cart. That filters the database naturally down to what's actually being eaten, and keeps it current the moment a manufacturer changes things.
Submitting itself is kept deliberately simple: two or three photos (front of pack, ingredients, nutrition table) are enough. We'll walk through the flow in a dedicated post once the beta is live.
Small is a feature, not a bug
We grow more slowly than any app that imports OpenFoodFacts, and we can live with that. A database where you search and find nothing useful isn't more helpful than one that returns plenty of hits – except those hits don't match reality. We'd rather show nothing than push an outdated entry at you with a "verified" stamp on it.
Without naming competitors: many apps chose the other path because big numbers sell better in the app store. We'll take the smaller number, and the work behind it.
Small here isn't a starting state we want to grow out of. Small stays small as long as "small" means: every product verified, every value traceable, every recipe change documented. We grow when you eat – and only then.