Methodology

Evidence collection

Read-only retrieval samples public HTML, Product and Offer structured data, canonicals, robots/sitemap availability, visible product facts, commerce actions and trust signals. Crawling is bounded by time, size and page limits.

Scoring

Five deterministic sub-scores cover Discoverability, Product Data, AI Readability, Commerce Readiness and Trust & Evidence. AI-assisted interpretation is constrained to collected evidence.

Evidence labels

FACT is directly retrieved data. OBSERVATION describes a deterministic check. INFERENCE is bounded interpretation. RECOMMENDATION is a proposed action.

Limits

Scores reflect a time-bound sample, not a complete crawl, security test or proprietary platform result. Sites can vary by location, cookies, JavaScript, login, bot controls and time.