How to Read Pet Food Comparison Tools: A Critical How-To Guide
Pet food comparison tools grade products from a printed ingredient list and composition, not from real digestibility, quality control or the maker's expertise (WSAVA, 2021). Their letter or point scores are reproducible but limited to what the label declares, and they vary from one site to the next. Read as one input among several, with the site's funding and method checked, they inform a decision; read as a verdict, they mislead.
Last updated :General documentary information. For an individual animal, a veterinarian's advice takes precedence over any online content.
What do rating-site scores actually measure?
Rating-site scores measure the printed composition and ingredient order through automatic rules, and little else. They cannot read real digestibility, finished-product control or whether a board-certified nutritionist was involved (WSAVA, 2021). A high score can therefore reward polished label-writing rather than nutritional validation, which is why it works only as a first read.
The mechanism is a fixed algorithm applied to declared data: the rank of the meat, an estimated carbohydrate level, the type and presence of additives, and free-from claims (WSAVA, 2021). Because these inputs all come from the bag, the output inherits the label's blind spots. The finding that undercuts the scores: two foods with near-identical ingredient lists can differ widely in real digestibility, a difference no label-based scale can detect (Tufts Petfoodology, 2023). The tool grades the description of a food, not the food.
This limitation is structural, not a flaw of any one site. No score can capture how well nutrients are absorbed, how tightly raw materials and finished product are tested, or who formulated the recipe, because none of that appears on the bag (WSAVA, 2021). The official AAFCO and FEDIAF nutrient profiles remain the only reference for adequacy. A score is best treated as a prompt to ask better questions, never as the answer.
How are the A to E grades calculated?
The A to E grades rest on an automatic scale applied to the label: the algorithm weights the ingredient order, an estimated carbohydrate level and the type of additives, then converts the result into a letter or point (WSAVA, 2021). The calculation is reproducible but bounded by what the label declares, and each site sets its own weightings.
Understanding the scale is the precondition for using it. The carbohydrate figure, for instance, is usually an estimate by difference, the nitrogen-free extract, since carbohydrate is rarely printed directly and is inferred from the other declared constituents (AAFCO, 2024). Ingredient order is taken at face value despite being ranked by weight as received, before cooking. The eye-opener is that the same food can swing from an excellent to a poor grade depending on how heavily one criterion is weighted, which means the letter describes the scale as much as the food.
What the scale cannot reach is the decisive part. It measures neither digestibility, nor raw-material quality, nor finished-product control, nor the maker's expertise, and it is not an official nutrient profile (Tufts Petfoodology, 2023; AAFCO, 2024). Before granting a single letter any weight, a reader benefits from finding the site's stated method: which criteria it scores, how it weights them, and how it treats grains and additives. A grade whose method is hidden cannot be interpreted at all.
Why does the same food score differently on different sites?
The same food scores differently because each site applies its own criteria, weightings and thresholds, and treats additives and grains in its own way (WSAVA, 2021). One scale can reward what another penalises, so an unchanged recipe earns an excellent grade in one place and a mediocre one in another. The product does not vary; the methodology does.
The divergence is most visible around contested choices. A site that penalises grains will downgrade a grain-inclusive food that a digestibility-focused site rates highly, working from the identical label (WSAVA, 2021). How carbohydrate is estimated, whether a meat meal is rewarded or treated with suspicion, and where the additive thresholds sit all push the letter up or down. The illustration that lays it bare: two scales can reach opposite verdicts on the same bag, which is only possible because neither is an objective measurement.
The lesson is to read contradiction as information. When grades conflict, that spread reveals a clash of methodological stances, not a fact about the food's quality (Tufts Petfoodology, 2023). The stable references do not shift with scoring fashion: the AAFCO and FEDIAF adequacy profiles, the maker's verifiable answers on formulation and control, and the food's fit to the individual animal (AAFCO, 2024; WSAVA, 2021). Returning to those when the letters disagree is the reliable move.
How can you tell if a tool earns money from affiliate links?
A tool's affiliate model shows in concrete clues: buy buttons that lead to retailers, tracking parameters in those links, mentions of affiliate, partner or promo codes, and a funding disclosure, often discreet, in the legal or about pages (Tufts Petfoodology, 2023). A complete absence of transparency about ownership and funding is itself a warning sign.
An affiliate site earns a commission on the sales it drives, which creates a structural incentive to tilt scores toward the products that pay (Tufts Petfoodology, 2023). In several countries a disclosure is legally required, so the words affiliate link or in partnership with frequently appear somewhere, even if buried. The point worth flagging: a site can display apparently neutral grades while monetising every click toward a seller, and it can look independent while being owned by a brand or retailer, with that fact tucked into the legal notices.
| Clue | Where to look | How to read it |
|---|---|---|
| Buy buttons to shops | Product pages | Affiliation likely |
| Tracking parameters in links | The redirect URL | Commission on clicks |
| "Affiliate" or "partner" notice | Legal or about pages | Disclosure present |
| Scoring method | About or methodology page | Transparent or opaque |
| Site ownership | Legal notices | Independent, brand or retailer |
A credible tool publishes its method, its sources and its business model in plain sight (Tufts Petfoodology, 2023). Where the model stays opaque, the score loses evidential value and should be cross-checked against non-commercial references before it influences a purchase.
Does a top-rated food automatically suit your own pet?
A top grade does not mean the food suits a particular animal. A score rates a recipe in the abstract on a label-based scale, knowing nothing of the animal's species, life stage, activity, neutering status or health, which are what actually decide suitability (WSAVA, 2021). An individual assessment cannot be replaced by a letter.
The gap between a general grade and a particular animal is easy to make concrete. A highly rated food that is rich and energy-dense can be wrong for a sedentary neutered [spayed] cat and feed its weight gain, while suiting an active dog perfectly well (NRC, 2006). The grade saw a sound recipe on paper; it could not see the cat. Needs vary sharply enough that the same food is right for one animal and unsuitable for another, which no score adjusts for.
Suitability is confirmed in the field, not on a screen. A food has to target the correct species and life stage, account for any medical need, and then prove itself on the animal over several weeks through body condition, stools and coat (WSAVA, 2021). For a sick animal a veterinary opinion outranks any ranking, however confident the letter looks. The score, at best, narrows the field; the animal settles the choice.
Can customer reviews and star ratings be trusted?
Customer reviews deserve distance. A star average mostly reflects palatability, convenience, price or delivery, and rarely real nutritional value (Tufts Petfoodology, 2023). It measures neither digestibility, nor quality control, nor adequacy, so it is a weak signal that should never be a main criterion for choosing a food.
The core problem is that an animal enjoying a food does not show the recipe is balanced. A highly palatable food can owe its appeal to added aromas with no link to its nutritional balance, so enthusiasm in the reviews says more about taste than about formulation (Tufts Petfoodology, 2023). Reviews are also vulnerable to solicited comments, fabricated entries and fashion effects, any of which can inflate or distort an average well away from the food's real merits.
Their one genuine use is pattern detection. A repeated complaint about batch-to-batch variability, packaging defects or frequent digestive upsets is worth noticing and verifying, because a recurring signal can point to a real quality-control issue (WSAVA, 2021). Even then, the signal is cross-checked against nutritional adequacy and the maker's seriousness, never accepted on its own. Read as a flag for further checking rather than as a score, reviews have a modest, defined place.
A side-by-side reliability table
| Tool | What it grades | Blind spots | Reliability alone |
|---|---|---|---|
| Letter / point rating sites | Printed composition, ingredient order | Digestibility, control, expertise | Low: first read only |
| A to E algorithmic scales | Label data, estimated carbohydrate | Method varies, no field test | Low |
| Affiliate comparison sites | Products, often with buy links | Commercial bias if undisclosed | Very low without transparency |
| Customer star reviews | Palatability, convenience, service | Adequacy, real quality | Very low; useful only for patterns |
| Official AAFCO / FEDIAF profiles | Nutrient adequacy | Not a quality ranking | High for adequacy |
| WSAVA manufacturer assessment | Formulation, trials, control | Needs the maker's cooperation | High |
The table sorts the tools by what they can and cannot see: the algorithmic and review-based tools sit low, while the institutional profile and the manufacturer assessment carry the weight.
A clear recommendation on using comparison tools
Use comparison tools as a first filter, never as a verdict. A score can shortlist foods and surface questions worth asking, but it grades a label and inherits the label's blind spots (WSAVA, 2021). Before trusting any letter, find the site's method and funding: a published methodology and a disclosed business model are the minimum for a score to mean anything, and undisclosed affiliate links are reason to discount it (Tufts Petfoodology, 2023).
When tools disagree, treat the contradiction as a signal to return to stable references rather than to pick the kinder grade. Verify nutritional adequacy against the AAFCO or FEDIAF profile, put the maker to the WSAVA questions on formulation, trials and quality control, and confirm fit on the animal over six to eight weeks (AAFCO, 2024; WSAVA, 2021). A score, a star average and an influencer's enthusiasm are at most weak inputs to cross-check; the adequacy statement, the maker's transparency and the animal's own response are what decide.
Related reading (Read Food)
- FAQ: Can you trust the scores from online kibble-rating sites?
- FAQ: How are the A to E grades on pet foods actually calculated?
- FAQ: How can you tell if a comparison site earns through affiliate links?
- Glossary: digestibility
- Glossary: carbohydrate calculation (NFE)
- Hub: Choosing and judging quality
Sources: WSAVA, Global Nutrition Guidelines (2021); Tufts Petfoodology (2023); AAFCO, Understanding Pet Food (2024); NRC, Nutrient Requirements of Dogs and Cats (2006).