We believe food information tools should be transparent. Here's exactly how IngredScan analyses products and generates scores — including our limitations.
Data Sources
When you scan a barcode, IngredScan retrieves product data from Open Food Facts, a free, open, collaborative food product database with over 3 million products worldwide (180,000+ UK products). This data is contributed by volunteers and manufacturers, which means:
- Coverage is extensive but not complete — some products may not be in the database
- Data accuracy depends on contributors — ingredient lists, nutrition info, and images are user-submitted
- Product formulations change — a product scanned today might have different ingredients than the database entry
We supplement Open Food Facts with reference data from the UK Food Standards Agency (FSA) for additive regulations, USDA FoodData Central for nutritional benchmarks, and publicly available supermarket product pages.
NOVA Classification
We determine a product's NOVA score (1-4) by analysing its ingredient list for marker ingredients characteristic of each processing level. Specifically:
- NOVA 4 markers: we check for ingredients like high-fructose corn syrup, hydrogenated oils, modified starches, protein isolates, maltodextrin, emulsifiers (e.g. E471, E433, E322 beyond natural lecithin), flavour enhancers (E621-E635), artificial sweeteners, and artificial colours
- NOVA 3: products with salt, sugar, or oil added to whole foods, without NOVA 4 markers
- NOVA 2: pure culinary ingredients (oils, sugar, flour, etc.)
- NOVA 1: unprocessed or minimally processed whole foods
When Open Food Facts already includes a NOVA classification, we use it as a starting point and validate it against our own analysis. In cases of disagreement, we err on the side of the more conservative (higher processing) classification.
Quality Score (0-10)
Our proprietary Quality Score provides a more nuanced assessment than NOVA alone. It starts at 10.0 and applies the following adjustments:
- -1.5 points per NOVA 4 additive detected (maximum penalty: -4.0 points). This captures the degree of ultra-processing — a product with one emulsifier is penalised less than one with three emulsifiers, flavour enhancers, and modified starch.
- -1.0 point if the Nutri-Score is D or E (poor overall nutritional profile)
- -0.5 points if the Nutri-Score is C
- -1.0 point if saturated fat exceeds 5g per 100g
- -0.5 points if sugar exceeds 10g per 100g
- -1.0 point if salt exceeds 0.6g per 100g (aligned with UK FSA traffic light thresholds)
- +0.5 points if the product has organic certification
The final score is clamped between 0.0 and 10.0. A score of 7+ generally indicates a product with minimal processing and decent nutritional profile. A score below 4 indicates significant concerns.
Additive Risk Ratings
Each additive in our database is classified as low, medium, or high risk based on:
- EU Regulation 1333/2008 — the European framework for permitted food additives, including maximum levels
- UK FSA guidance — specific UK regulatory positions and advisory statements
- Published peer-reviewed research — we review meta-analyses, systematic reviews, and large-scale studies published in reputable journals
These are relative classifications. "High risk" means an additive has accumulated more published concerns relative to others — it does not mean it is dangerous at the levels found in food. All permitted additives have passed regulatory safety assessments.
Limitations
We want to be upfront about what IngredScan cannot do:
- We are not a medical device. Our scores are informational tools, not medical advice.
- We cannot account for individual allergies or intolerances. Always read the label if you have specific dietary needs.
- Our data may be outdated. Product formulations change, and database updates lag behind.
- NOVA has limitations. It was designed as a research classification system, not a consumer tool. Some classifications are debatable (e.g. is sourdough with a small amount of ascorbic acid NOVA 3 or NOVA 4?).
- Our Quality Score is opinionated. Any single-number summary of food quality involves trade-offs and simplifications. We've chosen weights that we believe are sensible, but reasonable people could disagree.
Continuous Improvement
We regularly review and refine our scoring methodology based on new research, user feedback, and data quality improvements. When we make significant changes, we document them and notify users. If you spot an error in a product's data or score, use the Report button on the result page — we review every report and apply corrections within 7 days.
For full details, see our disclaimer. Questions about our methodology? Contact us at support@ingredscan.com.