Use this checklist as an audit record. For each item, save the evidence, apply the stated pass/fail condition, set severity, assign an owner, document the correction, and repeat the validation step after deployment. A check is not complete because the configuration exists in code; it passes only when the expected behavior is observable from the rendered page, crawl, response, or search diagnostic.
Rendering and indexability foundation - Critical
- Rendered primary content. Evidence: compare initial and rendered HTML for representative routes and confirm the visible service or project content is present. Pass: the intended content can be retrieved and rendered without relying on an interaction. Fail: the crawler-facing result is an empty shell, loading state, or materially incomplete page. Severity: Critical. Owner: frontend or platform engineer. Corrective action: make essential content reliably render for crawlable routes using the architecture appropriate to the application. Validation: repeat the render test and compare the output with the user-facing page.
- Route-level document metadata. Evidence: inspect the rendered title, description, canonical reference, and other required head elements on distinct routes. Pass: each indexable route exposes the intended route-specific values. Fail: values are missing, duplicated unintentionally, or inherited from the application shell. Severity: Critical. Owner: frontend engineer with SEO reviewer. Corrective action: move metadata generation into the route rendering path and remove conflicting defaults. Validation: recrawl representative routes and inspect the final rendered head.
- Crawlable route addresses. Evidence: copy representative internal links, request them directly, and inspect navigation markup. Pass: important content areas resolve at stable path-based addresses and are linked with standard href attributes. Fail: important states exist only behind fragments, click handlers, or transient application state. Severity: Critical. Owner: frontend or routing engineer. Corrective action: expose indexable content through durable routes and normal links. Validation: start from a crawlable entry page and confirm a crawler can discover each target route.
- Canonical consistency. Evidence: inspect canonical references on templates that can create duplicate or alternate addresses. Pass: the canonical points to the intended preferred address and does not accidentally consolidate unrelated pages. Fail: canonicals are missing where needed, point to the wrong route, or are generated from stale application state. Severity: Critical. Owner: frontend engineer with SEO reviewer. Corrective action: derive canonicals from the resolved route and documented duplicate policy. Validation: crawl affected templates and compare canonical targets against the expected URL set.
- Critical JavaScript dependency behavior. Evidence: test representative routes with network and console diagnostics while observing whether essential content still becomes available. Pass: required content renders without uncaught errors or blocked essential assets. Fail: a script failure prevents the route from producing its primary content. Severity: Critical. Owner: frontend engineer. Corrective action: repair failing dependencies, error handling, or render sequencing. Validation: rerun the same route tests and confirm the content and metadata remain available.
URL and metadata structure - High
- Path routing. Evidence: inspect production URLs and route configuration. Pass: indexable pages use stable path-based addresses. Fail: search-relevant pages depend on fragment routing or non-addressable states. Severity: High. Owner: frontend engineer. Corrective action: migrate search-relevant content to path routing with server support for direct requests. Validation: request deep links directly and confirm the correct page is returned.
- Unique titles and descriptions. Evidence: export rendered metadata for the crawlable route set. Pass: important routes have descriptive values that match their actual page purpose without unintended duplication. Fail: metadata is absent, generic, or repeated across materially different pages. Severity: High. Owner: SEO editor and frontend engineer. Corrective action: define template logic and route-specific copy. Validation: recrawl and review duplicates and missing fields.
- Social metadata consistency. Evidence: inspect Open Graph and social preview fields on representative routes. Pass: values reference the intended page title, description, image, and address. Fail: the application shell or an unrelated route supplies the values. Severity: Medium. Owner: frontend engineer or content owner. Corrective action: bind social fields to resolved route data. Validation: inspect rendered source and a preview debugger without treating social tags as a ranking requirement.
- robots.txt access rules. Evidence: review the live file and test important paths against its directives. Pass: intended public routes and required resources are not accidentally disallowed. Fail: a broad rule prevents crawling of pages or assets needed to render them. Severity: Critical when important pages are blocked, otherwise High. Owner: platform engineer with SEO reviewer. Corrective action: narrow or remove conflicting directives. Validation: retest representative URLs and monitor crawl diagnostics.
- Duplicate-address handling. Evidence: identify query variants, trailing variants, alternate hosts, and application-generated duplicates. Pass: duplicates are consolidated or excluded according to an explicit policy. Fail: multiple crawlable addresses compete for the same content without a preferred version. Severity: High. Owner: platform or frontend engineer. Corrective action: use routing, redirects, canonicals, or parameter controls that match the duplicate type. Validation: recrawl the duplicate set and confirm only the intended version remains indexable.
Performance and crawl efficiency - High to Medium
- Critical rendering speed. Evidence: measure representative templates with field data where available and controlled lab tests for diagnosis. The previously published checklist used a First Contentful Paint target under 1.5 seconds and a Time to Interactive target under 3.5 seconds; treat those values as historical operating thresholds rather than guarantees or current Google ranking requirements. Pass: no severe render delay prevents users or crawlers from reaching the primary content. Fail: main content is consistently delayed by avoidable application work. Severity: High when rendering is impaired, otherwise Medium. Owner: frontend performance engineer. Corrective action: reduce blocking work, ship less unnecessary JavaScript, and prioritize essential rendering. Validation: rerun the same test conditions and compare the affected template.
- Route-level code delivery. Evidence: inspect application bundles and network activity across representative routes. Pass: pages do not download unrelated application code before essential content can render. Fail: route entry requires large avoidable bundles or expensive initialization unrelated to the page. Severity: Medium. Owner: frontend engineer. Corrective action: split code at meaningful route or feature boundaries and defer nonessential modules. Validation: compare bundle and render behavior after the change.
- Media delivery. Evidence: inspect image dimensions, formats, transfer size, and loading behavior on agency portfolio and case-study templates. Pass: images are appropriately sized and noncritical media does not block primary content. Fail: oversized assets or eager loading materially delay the page. Severity: Medium. Owner: frontend engineer or design systems owner. Corrective action: resize, compress, use appropriate modern formats, and load offscreen media when useful. Validation: repeat network and rendering tests on the same route.
- Compression and production assets. Evidence: inspect response headers and delivered JavaScript or CSS. Pass: production assets are minified and transferred with appropriate compression supported by the stack. Fail: unminified or unnecessarily large assets increase transfer and parse cost. Severity: Medium. Owner: platform engineer. Corrective action: correct the production build and delivery configuration. Validation: request assets again and verify smaller optimized responses.
- Low-value URL discovery. Evidence: compare crawl data, internal links, sitemap entries, and Search Console examples for parameterized or duplicate states. Pass: search engines are guided toward canonical content rather than large sets of low-value variants. Fail: application filters, sort states, or generated URLs create avoidable discovery noise. Severity: High at scale, otherwise Medium. Owner: SEO reviewer with frontend or platform engineer. Corrective action: remove unnecessary crawlable links, consolidate duplicates, and keep sitemaps focused on canonical pages. Validation: recrawl and review whether the unwanted URL pattern declines in discovery.
Structured data, sitemaps, and discovery - High to Medium
- Relevant structured data. Evidence: inspect rendered JSON-LD on pages where a documented schema type accurately describes visible content. Pass: markup matches the page and contains supported properties sourced from the page itself. Fail: markup is missing where the project intentionally requires it, contradicts visible content, or describes entities that are not present. Severity: Medium. Owner: SEO reviewer and frontend engineer. Corrective action: add or correct only relevant markup. Validation: inspect the rendered document and run an appropriate validator; do not treat structured data as a guaranteed ranking factor.
- Structured data syntax. Evidence: run representative pages through Google's current validation tools where applicable and review parse errors. Pass: the markup parses and the values match the page. Fail: syntax errors or mismatched data prevent reliable interpretation. Severity: Medium. Owner: frontend engineer. Corrective action: fix serialization, escaping, property placement, or template logic. Validation: rerun the same validator and manually compare the output with visible content.
- XML sitemap coverage. Evidence: compare the submitted sitemap with the canonical indexable route inventory. Pass: important canonical pages are included and excluded routes are intentional. Fail: valuable routes are absent, obsolete routes remain, or noncanonical variants dominate the file. Severity: High when discovery is materially affected. Owner: platform engineer with SEO reviewer. Corrective action: generate the sitemap from canonical route data. Validation: fetch the live sitemap, sample URLs, and review Search Console processing.
- Paginated and segmented content. Evidence: inspect routes that spread meaningful content across multiple pages or states. Pass: each useful segment is reachable with standard links and has a clear canonical and indexing policy. Fail: later content is accessible only through scripts or the configuration creates duplicate crawl paths. Severity: Medium. Owner: frontend engineer and SEO reviewer. Corrective action: expose stable linked routes or simplify the experience where separate pages are not useful. Validation: crawl from the first segment and confirm intended discovery and canonical behavior.
- Internal link markup. Evidence: inspect navigation, cards, breadcrumbs, and contextual links in rendered HTML. Pass: crawlable destinations use standard anchor elements with href values. Fail: important navigation depends only on scripted click events. Severity: High for orphaning risk. Owner: frontend engineer. Corrective action: render real links while preserving application behavior. Validation: crawl without simulated user interactions and confirm the target routes are discovered.
Monitoring and release control - Ongoing
- Core Web Vitals monitoring. Evidence: review available field data in Search Console or another measurement source and pair it with lab diagnostics when investigating regressions. Pass: the team can identify affected templates and distinguish field evidence from test data. Fail: performance decisions are based only on anecdotal page loads. Severity: Medium. Owner: performance engineer or technical SEO owner. Corrective action: establish repeatable monitoring tied to route groups. Validation: confirm the reporting view contains the intended templates and can detect a known test change.
- Third-party script impact. Evidence: profile tag managers, chat widgets, analytics, embeds, and other external scripts. Pass: nonessential scripts do not materially delay primary content or destabilize rendering. Fail: third-party work blocks rendering or causes route errors. Severity: High when core content is affected, otherwise Medium. Owner: frontend engineer with marketing owner. Corrective action: defer, remove, or conditionally load nonessential scripts. Validation: rerun performance and rendering tests with the revised configuration.
- Crawl and indexing alerts. Evidence: define monitored Search Console properties and issue review responsibilities. Pass: the team has a repeatable way to notice material indexing or crawl changes after releases. Fail: regressions can persist unnoticed because no owner reviews diagnostics. Severity: Medium. Owner: technical SEO owner. Corrective action: add a release-linked review process and documented escalation path. Validation: confirm the responsible owner can retrieve affected examples and reproduce the check.
- Mobile rendering parity. Evidence: compare mobile and desktop rendered content and test interactive elements needed to reach crawlable pages. Pass: essential content, links, metadata, and status behavior remain consistent across supported layouts. Fail: responsive logic hides or replaces important content or navigation. Severity: High when discoverability changes. Owner: frontend engineer. Corrective action: fix responsive rendering and navigation logic. Validation: recrawl representative pages with mobile rendering and compare outputs.
- International targeting where applicable. Evidence: inspect alternate-language or regional pages and their hreflang relationships only when the site genuinely maintains such versions. Pass: reciprocal annotations use valid language or region codes and point to equivalent canonical pages. Fail: annotations are missing, contradictory, or map users to non-equivalent content. Severity: Medium. Owner: SEO reviewer and frontend engineer. Corrective action: generate annotations from the maintained locale map. Validation: crawl the international set and verify reciprocal relationships.
- Legacy or optional delivery variants. Evidence: inventory any alternate page technology or special delivery format still present in the application. Pass: each retained variant has a documented purpose and does not create conflicting indexable duplicates. Fail: obsolete or redundant variants remain discoverable without a clear canonical policy. Severity: Low to Medium. Owner: platform engineer. Corrective action: remove unused variants or align their indexing and canonical behavior. Validation: recrawl the affected URL pattern and confirm the intended version is discoverable.
- Post-release re-audit. Evidence: link major routing, rendering, metadata, or template releases to a focused crawl and Search Console review. Pass: changes that can affect search visibility are revalidated after deployment. Fail: the team assumes a successful application release also proves crawlability and indexing behavior. Severity: Ongoing governance. Owner: engineering lead and technical SEO owner. Corrective action: add search validation to the release checklist. Validation: repeat the relevant checks on production and record the result.