How does Ingredo.app score a product?
Ingredo.app builds a transparent base score from the ingredient list and source-cited data about each ingredient, then re-scores that same product against your own condition and diet profile, so the final verdict is personal, not generic.
The base score answers "is this generically good"; the personal verdict answers "is this good for me." Both are shown together, with the reasoning behind each visible, never a bare number.
Nothing in the pipeline is guessed: an ingredient we don't recognise is shown as unrecognised, never as a pass, and every flag traces back to a cited source.
The base score
For food, the score is mostly about the ingredient list. Ingredient quality is worth 50 points, weighted toward the first ingredients because EU labels list them by weight; processing level (NOVA) is worth 15, additives 15 and organic 10. Nutri-Score only nudges the result, by up to 10 points either way, and a single high-risk additive caps the score at 49. A product with no ingredient list is not rated at all — we show the Nutri-Score printed on the pack and ask you to scan the ingredients instead.
For cosmetics, every product starts at 100 and loses points for ingredients of concern: EU-labelled fragrance allergens (5 points each, up to 25), sulfate cleansers, parabens, known contact allergens, pore-clogging ingredients and similar. Each kind has a limit, so five allergens cost more than one but no single kind can sink a product on its own. A serious hazard — an isothiazolinone preservative (MI/MCI) or a formaldehyde releaser — caps the score at 20 or 40 whatever else is in it. Rinse-off products such as shampoo and shower gel count each concern at half weight; we tell from the product name or its ingredients, and treat a product we can't place as leave-on. A cosmetic is rated only once we recognise at least half its ingredients.
Allergies are handled by your profile, not by the general score. Fragrance allergens matter a great deal to people who react to them and very little to everyone else, so the base score takes off a few points while the Fragrance-allergy profile turns any EU fragrance allergen into a hard block.
Where an ingredient's classification is proposed by an offline AI enrichment step rather than a verified source, the score is marked as AI-influenced and the contributing estimates are listed for you to inspect.
Your personal verdict
Your profile — the condition and diet modules you've turned on — is layered on top of the base score. Each active module reacts to specific ingredient tags: gluten sources for Celiac, high-FODMAP ingredients for IBS, Annex III allergens for fragrance allergy, and so on.
Each ingredient tag maps to one of three flags: avoid (a hard block, like gluten for Celiac), caution (a soft penalty, like FODMAPs for IBS), or good match. Hard blocks come only from verified tags — our own source-cited knowledge base, or an AI proposal confirmed by at least three independent authoritative sources. An unverified AI estimate can never set one.
Common questions
Yes. Methodology changes are versioned, and a product's score can change if its recipe, the underlying data, or our ruleset is updated.
No. Offline AI enrichment only proposes tags for ingredients with insufficient data, and those proposals never produce a hard-block on their own — they go through human review or a deterministic cross-check first.