Use this section to decide whether the catalog can be crawled, rendered, and understood as intended. Sample the templates that generate the most product and collection URLs, and keep evidence from the live site rather than relying only on theme settings or development assumptions.
Checkpoint: Verify product-gallery image delivery. Evidence required: record the requested asset, rendered dimensions, transferred weight, format, priority, and loading behavior on representative product pages across common viewport classes.
Pass: the browser receives an appropriately sized asset, important product detail remains inspectable, layout stays stable, and the gallery does not create an avoidable delivery bottleneck. Fail: oversized source files are repeatedly delivered where responsive derivatives would suffice, visible assets arrive late because priority is misconfigured, or image behavior interferes with product interaction.
Severity: High when the pattern affects a shared product template. Owner: front-end or ecommerce developer. Corrective action: generate responsive derivatives, use modern formats where they preserve the required product detail, match loading priority to the image's actual role, and keep zoom or alternate-view needs intact.
The retained 5MB value is a source audit example, not a universal acceptance threshold. Validation: retest the same sample after release, confirm the network request now delivers the intended variant, and compare the rendered page in PageSpeed Insights or Lighthouse. Tools: Cloudinary, Squoosh, Next-gen Image Converters
Checkpoint: Control faceted-navigation crawl paths. Evidence required: crawl selectors for material, gemstone, style, price, size, availability, and other catalog filters; then record indexability, canonical targets, internal-link exposure, parameter behavior, and sitemap inclusion for representative states.
Pass: intentionally indexable states have stable URLs and distinct useful page purpose, while repetitive combinations consolidate consistently to the preferred destination and are not exposed as uncontrolled crawl paths.
Fail: selectors create large sets of duplicative URLs, canonicals conflict with indexability, internal links point broadly to states with no search purpose, or sitemap entries contradict the intended policy.
Severity: High for large or frequently refreshed catalogs. Owner: technical SEO lead and ecommerce developer. Corrective action: define the allowed indexable facet set, align canonical and internal-link rules, remove unnecessary crawlable combinations, and document any state that intentionally remains available to users without being a search landing page.
Validation: recrawl with parameter handling visible, inspect representative live URLs in Google Search Console, and confirm the resulting state matches the documented rule. Tools: Google Search Console, Screaming Frog
Checkpoint: Audit the primary mobile loading bottleneck on core templates. Evidence required: save the loading waterfall, identify the Largest Contentful Paint element, and compare discovery timing, transfer size, script work, and render dependencies for product, collection, and home templates.
Pass: the main above-the-fold asset or content is discovered promptly and delivered without an avoidable dependency that the template can remove. Fail: the principal content waits on oversized media, late script execution, unnecessary blocking resources, or a template pattern that delays discovery.
Severity: High when the same problem is repeated across a large page set. Owner: front-end developer. Corrective action: prioritize the actual critical resource, reduce avoidable blocking work, right-size media, and move noncritical behavior out of the critical loading path where the application permits.
The source used 4 seconds as prior diagnostic context; keep it as an example for investigation rather than a promise about bounce behavior or search outcomes. Validation: rerun the same template and device class in PageSpeed Insights and Lighthouse after deployment and compare the evidence before closing the issue. Tools: PageSpeed Insights, Lighthouse
Checkpoint: Find broken and wasteful internal navigation paths. Evidence required: crawl persistent menus, footer links, breadcrumbs, collection modules, product recommendations, editorial links, and other recurring navigation for error responses, unnecessary redirect chains, and obsolete destinations.
Pass: intended internal destinations resolve to the correct live page and recurring navigation does not send users or crawlers through broken or misleading paths. Fail: dead destinations remain linked from persistent components, redirect chains obscure the intended target, or product and collection modules point to pages that no longer serve the described purpose.
Severity: Medium to High based on prominence and scale. Owner: content operations or development team. Corrective action: update the link to the correct live destination, restore a page when the underlying customer need still exists, or remove the obsolete reference when there is no valid replacement.
Validation: recrawl the affected components and manually test the highest-visibility paths on the rendered site. Tools: Ahrefs, Sitebulb
Checkpoint: Validate Product structured data against the live product record. Evidence required: compare the JSON-LD on representative eligible product pages with visible product identity, price, availability, offer details, review information, and other supported properties that genuinely exist.
Pass: markup parses, describes the same product a customer sees, and contains only values the business can support from its catalog or customer-facing records. Fail: required information for the chosen markup is malformed, availability or price disagrees with the page, product identity is inconsistent, or unsupported claims are inserted solely for search presentation.
Severity: High when a shared implementation affects much of the catalog. Owner: ecommerce developer with SEO review. Corrective action: generate markup from the same reliable data source used by the product page, remove unsupported properties, and align variation handling with the visible experience.
Validation: test representative live URLs with Schema.org tooling and Google Rich Results Test, then inspect the rendered markup after deployment. Structured data can support machine understanding and eligibility for supported search features, but it does not guarantee review displays, shopping placement, clicks, or organic treatment. Tools: Schema.org, Google Rich Results Test