Technical SEO Audit Checklist (2026): The 80-Point Framework

The template approach in one paragraph

Most published technical SEO audit checklists are either too high-level to action ('check robots.txt') or too long to use ('487 items, with 300 redundant'). This is the working 80-point checklist we use on real Canadian client audits in 2026, organized into eight workstreams: crawlability and indexation, Core Web Vitals and performance, schema and structured data, mobile and accessibility, internal linking and IA, security and tech-debt, AI-search readiness, and Canadian-context items (hreflang, Markets/multi-region, accessibility law). Work through it linearly. Document each finding with severity and effort. The output is a prioritized backlog, not a 60-page PDF nobody reads.

How to use the checklist

Work through the workstreams linearly. For each item, document: status (pass/fail/partial), severity (P0 / P1 / P2 / P3), effort estimate (hours), and notes. The output should be a prioritized backlog of fixes — not a narrative report.

Tools you'll need: Screaming Frog or Sitebulb (site crawler), Google Search Console access, PageSpeed Insights, Schema.org validator, accessibility testing tool (axe DevTools or WAVE), and a representative mobile device (real device, not just emulator).

Crawlability and indexation (12 items)

• 1. robots.txt is correct and not blocking critical assets

  Pull /robots.txt. Verify CSS/JS aren't disallowed. Verify staging-environment Disallow: / hasn't leaked to production.
• 2. XML sitemap exists and is valid

  /sitemap.xml or /sitemap_index.xml. Validates against XML schema. Submitted in GSC.
• 3. Sitemap doesn't list noindex / 404 / redirected URLs

  Crawl the sitemap; flag anything that doesn't return 200 + index.
• 4. GSC Pages report: Indexed count is reasonable for site size

  If indexed count is dramatically lower than expected, investigate.
• 5. GSC Pages report: 'Crawled, not indexed' has a manageable count

  A bloated 'crawled not indexed' bucket suggests thin content or canonicalization issues.
• 6. No infinite-loop crawl traps

  Faceted URLs that generate combinations infinitely. Calendar widgets without crawl boundaries. Common in WooCommerce, Shopify with poorly-configured filters.
• 7. Canonical tags self-reference where appropriate

  Every indexable page canonicals to itself unless intentionally consolidating. Common bug: canonical pointing to root or homepage.
• 8. Pagination handled correctly

  If using paginated lists, decide between rel=next/prev (deprecated by Google but still respected by Bing), self-canonical with prevention of duplicate content, or noindex on pages 2+. Document the decision.
• 9. No accidental noindex on commercial pages

  Crawl and check meta robots / X-Robots-Tag headers on every commercial-intent URL.
• 10. Redirect chains under 2 hops

  Long redirect chains lose equity. Map and consolidate.
• 11. Internal links use canonical URLs (no parameters, trailing slash discipline)

  Inconsistent internal linking creates duplicate-URL signals.
• 12. 404s have helpful content + return correct status code

  Soft 404s (404 page that returns 200) confuse Google; verify the response code.

Core Web Vitals and performance (10 items)

• 13. CRUX field data passing on 75%+ of templates

  GSC Core Web Vitals report. Use field data, not lab data, as truth.
• 14. LCP < 2.5s on key templates

  Homepage, top 5 commercial templates, top 5 content templates.
• 15. INP < 200ms on key templates

  INP replaced FID in 2024. Audit interactive elements: chat widgets, A/B testing scripts, analytics tags.
• 16. CLS < 0.1 on key templates

  Reserved space for images, no late-loading ads/widgets that shift layout.
• 17. Image optimization: AVIF/WebP, properly sized, lazy-loaded

  Modern formats served via <picture>, srcset for responsive sizing, lazy-load below the fold.
• 18. Font loading: preload critical fonts, font-display: swap

  Prevents FOIT; reduces CLS from font swap.
• 19. CSS/JS bundle size: critical CSS inline, defer non-critical JS

  Lighthouse 'Eliminate render-blocking resources' should be empty.
• 20. Caching strategy: aggressive on static assets, sensible on HTML

  Cache-Control headers correctly set per asset type.
• 21. Third-party script audit: every tag justified, async/defer where possible

  Most sites carry 5–15 dead tags from old marketing tools. Audit and remove.
• 22. CDN edge presence in Canada (Toronto/Vancouver POPs)

  Cloudflare, Fastly, Akamai all have strong Canadian presence. Sites hosted US-East lose 80–200ms TTFB on Canadian traffic without a Canadian-edge CDN.

Schema and structured data (10 items)

• 23. Organization schema sitewide in <head>

  Single Organization @id used as reference everywhere else.
• 24. WebSite schema with SearchAction (if site has internal search)

  Enables Google sitelinks-search-box rendering.
• 25. BreadcrumbList schema on every nested page

  Improves SERP rendering, helps Google understand IA.
• 26. Article / NewsArticle / BlogPosting on content pages

  With author, datePublished, dateModified, image, publisher.
• 27. FAQPage on pages with genuine Q&A content

  5–10 real questions/answers. Don't FAQ-spam pages without real Q&A.
• 28. Product / Offer (with priceCurrency: 'CAD' for Canadian commerce)

  On every product page. AggregateRating where reviews exist.
• 29. LocalBusiness or specific subtype on local-services pages

  Use the most specific subtype (Dentist, AttorneyLaw, HVACBusiness, etc.) where available.
• 30. Author Person schema with @id

  Enables author-knowledge-panel features and improves E-E-A-T signaling.
• 31. Speakable schema on long-form content

  Enables voice-assistant excerpting. Low-effort, increasingly relevant for AI search.
• 32. All schema validates against Schema.org + Google Rich Results Test

  No syntax errors, no invalid types, no missing required properties.

Mobile and accessibility (10 items)

• 33. Mobile-first index status confirmed in GSC

  Settings → Indexing crawler. Should show 'Googlebot Smartphone'.
• 34. Tap targets ≥ 48×48 CSS pixels with adequate spacing

  Test on real device. Buttons/links too small or too close trigger Core Web Vitals 'tap target' warnings.
• 35. Text readable without zoom (16px minimum on mobile)

  Below 16px causes iOS to auto-zoom on focus, hurting UX.
• 36. Viewport meta tag correctly configured

  <meta name='viewport' content='width=device-width, initial-scale=1'>. Verify present sitewide.
• 37. Hamburger nav opens with adequate touch target and accessible focus

  Keyboard-accessible, screen-reader-announced.
• 38. Color contrast ratios meet WCAG 2.2 AA minimum

  4.5:1 for body text, 3:1 for large text. Test with axe DevTools.
• 39. All images have meaningful alt text (or empty alt for decorative)

  Audit at least the homepage and top 10 templates.
• 40. All form inputs have associated labels

  <label for=''> or aria-label. Critical for screen readers.
• 41. Heading hierarchy is logical (no skipped levels, one H1 per page)

  Affects both accessibility and SEO.
• 42. Site usable with keyboard alone (Tab, Enter, Escape)

  Nav, modals, forms all keyboard-operable. Visible focus indicators.

Internal linking and IA (10 items)

• 43. Every commercial page reachable in ≤3 clicks from homepage

  Pages buried 5+ clicks deep usually under-perform.
• 44. No orphaned indexable pages (zero internal links)

  Crawl identifies orphans. Either link them in or noindex them.
• 45. Anchor text is descriptive, not 'click here' / 'learn more'

  Audit top 100 internal links. Anchor text should describe the destination.
• 46. Top-level navigation: 4–7 items, not 12+

  More than ~7 nav items dilutes signal flow.
• 47. Footer links: include orphan-prone deep pages, but limit to 30–60 total

  Footer is a useful link surface, not infinite.
• 48. Hub pages exist for major topic clusters

  Each topical cluster has a hub page that links out to the cluster pages and back.
• 49. Related-content modules on long-form pages

  3–5 related pages per long-form article, contextually relevant.
• 50. Internal links to canonical URL versions only

  No links to /page?utm=internal or trailing-slash inconsistencies.
• 51. No-follow only used intentionally (paid links, untrusted UGC)

  Audit nofollow usage; remove from internal links.
• 52. Page depth distribution is reasonable

  Most pages within 3 clicks of homepage; nothing unintentionally beyond 5.

Security and tech-debt (8 items)

• 53. HTTPS sitewide, HTTP redirects to HTTPS

  Test with non-HTTPS request — should 301 to HTTPS.
• 54. SSL certificate valid and renews automatically

  Lapse-prone SSL is one of the most common avoidable downtime causes.
• 55. HSTS header set with reasonable max-age

  Strict-Transport-Security header. Improves security posture.
• 56. No mixed-content warnings

  All assets (images, scripts, fonts) loaded over HTTPS.
• 57. CMS, plugins, server software up to date

  Outdated WordPress, Shopify themes, etc. are both security and ranking risks.
• 58. No leaked staging environments

  Crawl for staging.* subdomains, dev.* subdomains, or preview branches that are accidentally indexed.
• 59. Server response codes correct

  200 for content, 301 for permanent redirects, 404 for missing, 410 for permanently gone, 503 for temporary outage. Wrong codes cause indexation issues.
• 60. CDN/server logs available for at least 30 days

  Required for diagnosing intermittent issues.

AI-search readiness (10 items)

• 61. /llms.txt published and current

  Top-level summary file for AI assistants. Lists key URLs and topics.
• 62. /llms-full.txt published with deeper content surface

  Full content listing in AI-friendly format.
• 63. Question-format article titles where intent matches

  AI assistants extract more readily from question titles.
• 64. Direct answer in the first paragraph of long-form content

  AI assistants extract lead paragraphs for citation.
• 65. FAQPage schema with real, substantive Q&A

  Highest-leverage schema for AI-citation visibility.
• 66. Author bio with credentials and Person schema

  AI assistants weight author identity heavily.
• 67. Updated dates visible and accurate

  Both human-readable and dateModified in schema.
• 68. Content not blocked from AI crawlers (unless intentional)

  Verify GPTBot, ClaudeBot, PerplexityBot are allowed in robots.txt unless you've made an intentional decision to block them.
• 69. Original perspective or data in cornerstone content

  AI-citation models prefer content with unique signal over content that's a synthesis of competitors.
• 70. Speakable schema on key content

  Improves voice-assistant excerpting, increasingly cited by AI surfaces.

Canadian-context items (10 items)

• 71. CAD currency in Product/Offer schema for Canadian commerce

  priceCurrency: 'CAD'. Critical for Canadian Shopping/SERP rendering.
• 72. addressCountry: 'CA' in LocalBusiness/Organization schema

  Establishes Canadian entity.
• 73. Canadian phone number format (+1)

  telephone: '+1-416-555-0100'. International prefix included.
• 74. Hreflang for English/French Canada (if bilingual)

  <link rel='alternate' hreflang='en-CA' href='...'> and 'fr-CA'. Bidirectional, correct, complete.
• 75. French content for Quebec audience (if serving Quebec commercially)

  Quebec Bill 96 imposes language requirements on commercial communications.
• 76. Canadian-hosted or Canadian-edge CDN for Canadian traffic

  Materially improves TTFB for Canadian users.
• 77. AODA accessibility compliance (if Ontario-serving)

  Ontario's Accessibility for Ontarians with Disabilities Act applies to most public-facing sites.
• 78. PIPEDA-compliant lead forms

  Privacy compliance for Canadian lead capture.
• 79. CASL-compliant email capture

  Marketing-email opt-in language and consent record-keeping.
• 80. GBP for Canadian locations with correct categories and Canadian address format

  Postal code format (A1A 1A1), province abbreviation (ON, BC, etc.).

Frequently Asked Questions

Implementation questions from Canadian operators.