Storefront SEO audit with Vendure

Audit rendered product and collection pages. Give the user a prioritized list of missing or incorrect SEO features and explain how each should be implemented using Vendure data. Do not modify code unless separately requested.

Ask first

• Which URLs, storefront/channel, locales, and environment should be audited?
• Is noindex intentional for this environment? Never remove it without approval.
• Does the user want an audit only or implementation afterward?

Audit

1. Inspect the rendered HTML, not only source components.
2. Record HTTP status and validate:
   • One descriptive <title> and one H1.
   • Unique meta description.
   • Absolute self-referencing canonical URL.
   • <html lang> and reciprocal localized hreflang, including x-default.
   • Intentional robots directive; flag noindex as the likely cause of a Lighthouse SEO drop.
   • Open Graph: title, description, image, type, URL, locale, site name.
   • Twitter: card, title, description, image.
   • Descriptive image alt text rather than asset filenames; intrinsic image dimensions.
3. Parse every application/ld+json block and validate page-specific schema:
   • Product: Product, Offer or AggregateOffer, BreadcrumbList; optionally AggregateRating, displayed Review, and FAQPage.
   • Collection: CollectionPage, BreadcrumbList, and ItemList.
4. Ensure schema matches visible/current data. Omit unknown values; never invent prices, stock, ratings, reviews, FAQs, or identifiers.
5. Report findings as concise bullets grouped by priority:
   • Blocking: indexing, status, canonical, invalid schema.
   • Important: missing title/description, hreflang, Product/Offer data.
   • Enhancement: social cards, richer ItemList data, alt text, dimensions.
6. Check /robots.txt for intentional crawl rules and an absolute sitemap URL. Check /sitemap.xml for complete, canonical, indexable product and collection URLs.
7. For every missing item, state both what is missing and how to implement it using the integration guidance below.

Vendure integration

• Add nullable, public metaTitle (localeString) and metaDescription (localeText) custom fields to both Product and Collection when missing. Group them under the SEO tab in the editor. Generate and run a database migration before saving values.
• Query localized product fields: name, slug, description, translations, featuredAsset, assets, collections, variants, and SEO custom fields such as metaTitle and metaDescription.
• Query variant sku, currencyCode, configured price (price or priceWithTax), and stockLevel for offers. Respect the storefront’s tax-display configuration.
• Build product metadata from trimmed, localized SEO custom fields. Fall back to trimmed product name and sanitized/truncated description. Render metadata through the shared page layout.
• Build canonical and hreflang URLs from localized translation slugs and the configured public site URL; do not use the request host on local/staging environments.
• Build Product schema from product/variant data, including SKU and images. Use Offer for a single variant and AggregateOffer for multiple variants, with low/high prices, currency, offer count, and availability.
• Convert Vendure minor-unit prices to major currency units before formatting offers (divide by 100 for two-decimal currencies). Map OUT_OF_STOCK to https://schema.org/OutOfStock and available stock levels to https://schema.org/InStock; ensure availability accurately reflects the variants.
• Add AggregateRating and Review only from the actual review source. Include only reviews represented by initial page data and containing valid author, text, and rating.
• Parse localized FAQ custom-field content defensively. Add FAQPage only for valid, visible question/answer entries.
• Build BreadcrumbList from collection breadcrumbs, including home and the product where applicable. Use absolute URLs consistent with canonical URL conventions and keep the trail synchronized with the visible breadcrumb component.
• Query collection name, slug, description, translations, breadcrumbs, featured asset, SEO custom fields, and products.
• Build collection metadata from trimmed, localized SEO custom fields, falling back to collection name and sanitized/truncated description. Render metadata through the shared page layout.
• Build CollectionPage schema containing an ItemList. Include position, absolute product URL, trimmed product name, and image when available. Deduplicate product-variant results by product ID and number positions starting at 1.
• Serialize structured data into application/ld+json scripts using the storefront framework’s safe serialization facilities.
• Query id, name, and preview for featured assets and asset lists, including assets nested under variants, order lines, and collections. Use an editor-maintained descriptive Asset.name for image alt text when available; otherwise fall back to the product name. Do not use generated filenames.
• Serve /robots.txt with the intended crawl rules and an absolute sitemap URL. Exclude search/filter crawler traps where appropriate and preserve intentional environment-specific indexing restrictions.
• Generate /sitemap.xml from paginated product and collection queries. Fetch slug, updatedAt, and totalItems; continue until all results are included. Include only canonical, indexable URLs with last-modified dates, using translated slugs for localized URLs. Choose static generation, an endpoint, or caching to match the storefront framework.
• Verify ownership in Google Search Console and Bing Webmaster Tools, then submit the sitemap to both. Review crawl errors, indexing status, structured-data warnings, and Core Web Vitals after deployment.

Verify

• Metadata values are present, unique, localized, and use absolute production URLs.
• Canonical and hreflang URLs use translated entity slugs and agree with navigation URLs.
• JSON-LD parses successfully and contains only supported, truthful fields.
• Product prices use major currency units, not Vendure minor units.
• Availability accurately summarizes variants.
• Image alt text uses descriptive asset names or product names, not generated filenames.
• Robots rules preserve the intended indexing policy and reference the absolute sitemap URL.
• The sitemap includes all canonical, indexable product and collection URLs with translated slugs and last-modified dates.
• A production build and type check pass; report unrelated pre-existing failures separately.
• Optionally validate JSON-LD with Google Rich Results Test or Schema.org Validator; eligibility does not guarantee rich results.

Limitations

• noindex intentionally retained by the user will continue to reduce Lighthouse SEO and prevent indexing despite other improvements.
• Storefront fallbacks cannot repair weak or missing localized content in Vendure.
• Google may ignore valid review or FAQ schema and applies additional eligibility policies.