Ecommerce Technical Seo Audit

Ecommerce Technical SEO Audit

Sample a handful of category, collection, and product pages to check for the technical-SEO problems that are specific to ecommerce catalogs rather than generic site-wide issues.

Scope: this is a spot-check for free-tier credibility-building, not an exhaustive audit. Sample 3-5 category/product pages and report what the sample actually shows — this is not a full-catalog crawl, and the report must never claim or imply the entire catalog was checked. The exhaustive, full-catalog, traffic-prioritized version of these same checks (crawling every category and product URL, cross-referenced against real GA4 traffic/conversion data) lives in a separate private/paid pack (verified-audit-skills's 04-ecommerce-verified-audit-wrapper) — this skill is the free teaser, not a duplicate of that paid differentiator.

Generic sitemap coverage, canonical tags, and broken-link sweeps belong to sitemap-discovery-audit; crawler/robots access rules belong to robots-ai-crawler-audit; structured data depth (Product, Offer schema) belongs to schema-markup-audit; whole-site triage belongs to ai-visibility-audit. This skill only covers the four catalog-specific checks below.

Workflow

1. Sample 3-5 category/collection pages and 3-5 product pages — pick a mix (a top-level category, a narrow sub-category, a well-known product, and, if discoverable, an older/less prominent one). Do not attempt to crawl the full catalog.
2. Check each sampled category/collection page for thin content: word count of unique on-page text (excluding nav, product-grid text, and boilerplate) and whether it duplicates another category's copy.
3. Check for faceted-navigation duplicate URLs: apply a filter or sort option on a sampled category page and check whether it produces a separately crawlable/indexable URL (query parameters, no canonical back to the base URL, no noindex, not blocked in robots.txt).
4. Check for orphan pages: for each sampled URL, check whether it is reachable from on-site navigation (nav, category grid, internal links) as well as present in the sitemap — a page in the sitemap with no discoverable internal link pointing to it is an orphan.
5. Check discontinued-product handling: if an old/discontinued product URL is discoverable (from the sitemap, a search engine cache, or a user-supplied list), check its status code and whether it redirects to a relevant replacement product/category or returns a proper 404/410 — as opposed to a generic 404 with no path forward, or a soft-404 (200 status with error-page content), or a redirect to an unrelated page/homepage.
6. Report findings as a sample, explicitly stating the sample size and pages checked — never as a catalog-wide conclusion.

Use the commands in references/checks.md for each check.

Output

• Sample scope (exact pages/URLs checked, and that this is a sample, not the full catalog)
• Thin category/collection page findings (word count, duplication)
• Faceted-navigation duplicate URL findings
• Orphan page findings
• Discontinued-product handling findings
• Recommended fixes
• Verification commands

Guardrails

See Shared Guardrails for the cross-cutting rules on not claiming AI platform outcome guarantees, not fabricating unverifiable claims, and not exposing private/sensitive paths.

• Never state or imply that the full catalog, or "all products/categories," was checked — state the exact sample size and which pages were sampled every time.
• Do not extrapolate a sample finding into a catalog-wide count (e.g. do not claim "N thin pages site-wide" from a 3-5 page sample) — report what the sample shows and recommend a full-catalog crawl as the next step if the client wants exhaustive coverage.
• Do not recommend a specific word-count target as a ranking guarantee — thin-content thresholds are practitioner guidance, not an official Google minimum; say so.
• Do not recommend redirecting a discontinued product to an irrelevant page (homepage, an unrelated category) — an irrelevant-target redirect risks being treated as a soft 404.