What it covers
Crawling — whether engines can reach pages.robots.txt rules, internal linking, crawl budget, redirect chains, and server response codes.
Indexation — whether crawled pages are eligible to appear. noindex directives, canonical tags, duplicate content, thin pages, and parameter handling.
Rendering — whether the content engines see matches what users see. Client-side rendering can hide body text from crawlers that do not execute JavaScript, including many AI retrieval crawlers.
Architecture — how pages relate. URL structure, category depth, internal link distribution, pagination, and faceted navigation.
Structured data — machine-readable markup describing page content, usually Schema.org vocabulary in JSON-LD. Required for rich results and increasingly consumed by AI systems and shopping agents.
Performance — load and interaction speed, measured by Core Web Vitals.
Internationalization — hreflang annotations, country and language targeting, and avoiding duplicate content across locales.
Common failure modes on ecommerce sites
- Faceted navigation generating unbounded crawlable URL combinations
- Product variants split across near-identical pages without canonicals
- Out-of-stock products returning soft 404s or being removed without redirects
- Pagination that blocks crawlers from reaching deep catalogue pages
- Structured data that contradicts the visible page — most often price or availability
- Migrations that drop redirects and lose accumulated link and citation history