Our Methodology
Evidence-aware. Source-backed. Transparent.
Decision support for better pet choices.
Our methodology explains how iPickPet presents source-backed facts, iPickPet Score, Match Score, Data Confidence, ingredient interpretation, and review-first scanner output. Eligible dog and cat food details pair a compact iPickPet Score with Data Confidence and inspectable reasons; supplement details remain facts-first.
Food product cards may show the same context-free score, but listings never sort by it. Match Score, Ask scoring, supplement scoring, and Compare scoring stay off, and Compare never forces a single pick.
No evidence_level is populated, no medical claims are made, and no reviewRating, ratingValue, or score-derived rating schema is published.
Evidence-First
We prioritize verified sources and transparent criteria.
Explainable
You can inspect the factors behind active decision-support outputs.
Pet-First
Every decision supports pet wellbeing over marketing.
Runtime Status
Runtime Status (July 16, 2026)
Active now
- - Eligible dog and cat food details show compact iPickPet Score with Data Confidence
- - Eligible food cards show the same context-free iPickPet Score without score ordering
- - Methodology v1.3 factors and eligibility boundaries documented on this page
- - Label completeness / missing-value behavior
- - Ingredient Classification boundaries
- - Scanner Ingredient Review (review-first)
- - Evidence-aware readiness planning for explanation review
Approval-gated
- - Any expansion into supplements, Match Score, Ask, Compare, score sorting, or score-derived schema
Not active yet
- - Public supplement scoring
- - Public Match Score
- - Ask scoring
- - Score sorting
- - Single-product pick logic on compare
- - Rating or review schema
Scoring framework
iPickPet Score
A source-backed 0–100 score for label quality, transparency, and formulation signals.
Match Score
A separate context-fit score. The same product can match one need better than another.
Data Confidence
A smaller trust signal showing how complete and source-backed the recorded data is.
Our Guiding Principles
Independent
Not owned by pet food brands or retailers.
Transparent
Data sources and methodology boundaries are visible.
No Hidden Agendas
No promotional products in ranking logic.
Continuous
Methodology records and public wording update as runtime capabilities evolve.
Expert-Informed
Guided by veterinary nutrition expertise.
Balanced
Tradeoffs include benefits, risks, and context.
What is active in methodology today
iPickPet Score v1.3
Eligible dog and cat food details and cards show its compact numeric output. Unknown guardrails and malformed provenance earn no positive credit.
Data Confidence
Shown beside eligible food-detail scores as source completeness, never blended into product quality or used to order cards.
Score explanations
Eligible food details expose a compact, expandable explanation drawn from the same deterministic factors as the displayed score.
Evidence-Aware Readiness
Reviewed evidence and regulatory context can support explanation review, but evidence_level is not populated and evidence claims do not determine product superiority.
Label Completeness
Missing-value behavior helps interpretation of coverage gaps. Completeness does not equal correctness or quality.
Ingredient Classification
Educational interpretation boundaries are context-aware. Unresolved ingredients remain reviewable instead of forced into certainty.
Scanner Ingredient Review
Review-first workflow: OCR/manual text is reviewable and not verified truth. Helpful/Caution/Unknown are educational categories, not medical verdicts.
What feeds iPickPet Score today
Food products currently use deterministic label and formulation inputs such as named protein specificity, ingredient transparency, generic ingredient penalties, life-stage and adequacy statement presence, guaranteed-analysis completeness, calories, source URL, verified date, and positive or negative label signals. Supplement entries use their own source-backed label-transparency lane built around species, category, dosage, ingredient detail, provenance, and caution notes.
Match Score remains context-specific and is not public. Data Confidence stays a separate completeness measure shown with eligible food-detail scores. No score sorting, winner logic, Ask scoring, reviewRating, ratingValue, or AggregateRating schema is published.
We document sources, confidence boundaries, and methodology status so you can inspect what is active now, what is internal only, and what is still pending.
See Our Data & SourcesData sources and references
iPickPet combines official brand pages, recognized nutrient-profile frameworks, reviewed evidence fixtures, and transparent methodology records. AAFCO adequacy and life-stage claims support label context, not approval, endorsement, superiority, or individual suitability.
Evidence claims support explanation review and do not by themselves rank, rate, diagnose, prove product quality, or populate rating or evidence_level fields.