Snuffo derives a 0–100 score from declared pet-food label values, compares them with published nutritional references for species and life stage, and displays a separate confidence level for the quality and completeness of the underlying data.
What the score measures
Snuffo evaluates nutritional adequacy and balance from information available on the product label. It is designed to answer: given what the label declares, how does this product compare with the published references and the current Snuffo ruleset?
A score is Snuffo’s informational opinion based on disclosed facts and rules. It does not certify manufacturing quality, prove ingredient sourcing, predict digestibility, or decide whether a food is appropriate for a medical condition.
Popular opinion often calls a food “junk” from its ingredient list - named meat versus “by-products,” or the presence of grains. Snuffo deliberately does not score this way. An ingredient’s name does not establish its nutritional quality, digestibility, or value, and leading veterinary nutrition bodies regard the ingredient list as one of the least reliable ways to judge a food. Snuffo scores the nutrition a label declares against published references, and stays explicit about what a label cannot reveal.
Facts in, judgements separate
Inputs can include protein, fat, fibre, ash, moisture, energy, species, life stage, and whether the food is complete or complementary. The engine converts declared values into comparable dry-matter and energy-normalised measures where the data allows.
Published nutritional minima and legal reference values are kept separately from Snuffo’s scoring choices. Weightings, grade bands, curve shapes, and penalties are explicit ruleset judgements. This separation makes it possible to update a source fact without quietly changing the opinion layer - or change a judgement without presenting it as a scientific fact.
How the 0–100 score is built
The current model combines assessable subscores for protein adequacy, fat balance, carbohydrate load, and mineral load. A potential protein-quality lever is omitted when ingredient evidence is not sufficient; its weight is removed instead of treating missing evidence as a pass. Protein quality here means how digestible a protein is and whether it supplies the amino acids a pet needs - properties a label cannot establish reliably - so Snuffo declines to estimate it rather than guess from ingredient names.
- Protein adequacy compares protein per unit of energy with the relevant species and life-stage minimum.
- Fat balance rewards adequate fat while accounting for very high levels.
- Carbohydrate load uses calculated nitrogen-free extract and species-specific bands.
- Mineral load uses declared ash as a limited proxy for total mineral content.
Adequacy factors reduce the result where an assessable essential minimum is missed. Safety gates can cap a result when a supported critical threshold is triggered. Unassessable levers do not receive invented values.
Confidence is not the score
A high score and low confidence mean different things. Confidence describes how complete, direct, and verifiable the product data is. It can fall when values are missing or estimated, when only one source is available, when the food is complementary, or when the label cannot support a particular assessment.
Snuffo shows the confidence beside the score and explains the factors behind it. More reliable source data can change confidence without changing the ruleset.
Grade bands
| Grade | Score | Plain-language reading |
|---|---|---|
| A | 80–100 | Strong result on the assessable levers |
| B | 65–79 | Solid, with one or more areas to examine |
| C | 50–64 | Mixed result on the measured factors |
| D | 35–49 | Below the current references on important levers |
| E | 0–34 | Significant shortfalls or a supported safety cap |
The grade is a compact summary, not a feeding instruction. The numeric score, confidence, pet context, and detailed breakdown carry the useful nuance.
Current limits
- Most packaging does not disclose a complete micronutrient profile.
- Ingredient lists do not reliably prove ingredient quality, sourcing, processing, or bioavailability.
- Snuffo does not audit factories, suppliers, feeding trials, or batch-level quality control.
- Individual health conditions, medicines, allergies, and veterinary diets need professional guidance.
- Manufacturers can change recipes and packaging before a catalog record is updated.
Snuffo is general information, not veterinary advice. A product can score well and still be wrong for an individual animal.
Core references
- FEDIAF Nutritional Guidelines for Complete and Complementary Pet Food for Cats and Dogs (2025)
- AAFCO Official Publication (2014)
- National Research Council, Nutrient Requirements of Dogs and Cats (2006)
- Relevant European Union feed and additive regulations for supported safety checks
References supply facts and thresholds; they do not endorse Snuffo’s weighting or grade choices.
Corrections and versioning
Catalog data and the ruleset are versioned so a result can be traced to the information and rules used at the time. When better label evidence arrives, Snuffo can correct a product record and rebuild scores without changing the mobile app’s scoring logic.
To challenge a label value, source, or product match, use the in-app report action or email the barcode, country, current package evidence, and the issue to contact@snuffo.app.
Want the long version?
The full article documents the current formula, levers, confidence model, references, and limitations in more detail.