Why a 47-point checklist (and not 10, or 200)
Forty-seven is not a magic number. It is the number of distinct checks we have found materially affect Toronto-area SEO outcomes in 2026, after running this same framework across hundreds of sites in legal, dental, medical, financial, B2B SaaS, ecommerce, and local-services verticals. A ten-point checklist misses the discipline that compounds over quarters; a two-hundred-point checklist drowns the genuinely important findings in noise. Forty-seven is what fits, end to end, into a credible written audit a senior strategist can walk a buyer through in a single conversation.
An audit is not a checklist; it is a prioritised remediation roadmap. The forty-seven checks below are inputs. The output is the order in which we recommend you fix things over the next ninety days.
Companion to our free Toronto SEO audit framework and the LLM citation tactics work this audit informs.
Crawl & indexation (1–8)
- 1. Crawl coverage in Search ConsoleWhat percentage of submitted URLs are indexed? Anything below 80% on a healthy site warrants investigation. For ecommerce or programmatic sites, the threshold is lower (60–70%) but should be intentional — every excluded URL should have a documented reason.
- 2. Robots.txt reviewAre critical sections accidentally disallowed? Are crawl directives consistent with intent? We see at least one robots.txt error blocking valuable pages on roughly one in three audits.
- 3. XML sitemap hygieneSitemap submitted, current, returning 200, no 404s or noindex pages included, lastmod dates accurate. Sitemaps with stale lastmods get downweighted by crawlers.
- 4. Canonical tag auditAre canonicals self-referential where they should be, and pointing to the correct master URL where consolidation is needed? Mis-set canonicals are the most common cause of missing-page complaints.
- 5. HTTP status sweepTop 100 most-trafficked URLs returning 200; redirect chains under three hops; no soft-404s on important pages. We export the full status profile and flag anything anomalous.
- 6. Pagination handlingrel=prev/next is deprecated, but pagination still needs sensible canonical strategy and link discoverability. Common failure mode: page 2+ of category listings becoming orphaned.
- 7. Faceted navigation controlOn ecommerce and large content sites: are facet combinations under explicit indexation control? Uncontrolled facets create thousands of low-quality URLs that dilute crawl budget.
- 8. JavaScript renderingIf the site is JS-heavy, does Googlebot's rendered HTML contain the actual content and links, or only the unrendered shell? Check via URL Inspection tool. JS rendering issues are the most common silent SEO failure on modern Toronto-built sites.
Technical & Core Web Vitals (9–16)
- 9. LCP (Largest Contentful Paint)Below 2.5 seconds on mobile for 75th-percentile users. The hero image or above-fold text element is almost always the culprit when LCP fails.
- 10. INP (Interaction to Next Paint)Below 200ms on mobile. Replaced FID in March 2024 and is meaningfully harder to pass — heavy JavaScript on interaction is the typical failure mode.
- 11. CLS (Cumulative Layout Shift)Below 0.1. Reserved image dimensions, font display optimisation, and ad-slot sizing are the usual culprits when CLS fails.
- 12. Mobile usabilityTap targets, font sizes, viewport configuration. Fewer issues than five years ago but still occasionally broken on legacy sites.
- 13. HTTPS and mixed-content auditAll URLs HTTPS, no mixed-content warnings, HSTS header present, certificate valid and not nearing expiry.
- 14. Image optimisationModern formats (WebP/AVIF), responsive srcset, lazy loading on below-fold images, dimensions specified to prevent CLS. Image bloat is the most common Core Web Vitals failure on Toronto SMB sites.
- 15. Server response timesTTFB under 600ms at the 75th percentile. Slow server response is often the underlying cause of LCP failures and is much harder to fix than image optimisation.
- 16. CDN and caching strategyStatic assets cached aggressively; HTML cached appropriately for the site's update cadence; geographic distribution sensible for the brand's audience footprint.
On-page & content (17–24)
- 17. Title tag audit (top 100 URLs)Unique, under 60 characters, keyword-aligned, brand suffix consistent. Duplicate or templated titles are the most common on-page SEO failure on Toronto-built sites.
- 18. Meta description auditUnique, under 160 characters, written for click-through rather than for keyword stuffing. Meta descriptions are no longer a ranking signal but materially affect CTR from SERPs and AI Overviews.
- 19. H1 hierarchyOne H1 per page, sensible H2/H3 structure, no header-tag styling abuse. Header hierarchy matters more for AI engines parsing pages than for classical Google ranking.
- 20. Content depth on money pagesMoney pages (services, locations, vertical landing pages) at credible word count and topical depth for the vertical. Thin money pages are the most common reason serious Toronto verticals underperform.
- 21. Keyword cannibalisation reviewMultiple pages targeting the same primary keyword? Consolidation candidates identified and prioritised. Cannibalisation is silent but compounds over time.
- 22. Content freshness auditWhen were the top 50 pages last meaningfully updated? Stale content underperforms in 2026 SERPs and is rarely cited by AI engines.
- 23. E-E-A-T signal auditAuthor bios, credentials, citations, sourced statistics, fact-check dates. Particularly important for YMYL verticals (legal, medical, financial) where E-E-A-T directly affects ranking.
- 24. Brand mention consistencyBrand name spelled and styled consistently across the site, in title tags, in copy, and in schema. Inconsistent brand mention dilutes entity recognition by both Google and LLMs.
Schema & structured data (25–30)
- 25. Organization schema (root)Complete Organization or LocalBusiness schema on the homepage with name, URL, logo, address, contact points, sameAs to social and authority profiles. The root schema anchors the entity graph LLMs parse.
- 26. Page-type schema coverageArticle schema on blog and editorial pages; Service schema on service pages; FAQPage schema where appropriate; HowTo schema for how-to content; Product schema on commerce pages.
- 27. Author schema with credentialsPerson schema for authors with credentials, sameAs to authoritative profiles (LinkedIn, professional bodies, publications). Particularly important for E-E-A-T verticals.
- 28. Schema validationEvery schema block validates in Schema Markup Validator and Google's Rich Results Test. Invalid schema is worse than missing schema.
- 29. Schema graph cohesionSchema entities reference each other via @id where appropriate. A cohesive entity graph is materially more useful to LLMs than disconnected schema blocks.
- 30. AEO-specific schemaFAQPage and Article schema with quotable Q&A pairs, designed for engine extraction. See our <a href="/blog/schema-for-ai-engines-deep-dive" class="text-primary hover:underline">schema for AI engines deep dive</a> for the full pattern.
Internal linking & site architecture (31–34)
- 31. Internal link distributionMoney pages receiving credible internal link equity from contextual editorial links, not just from nav and footer. Audit via Screaming Frog or similar — the distribution rarely matches what the team expects.
- 32. Anchor text varietyInternal anchors varied and descriptive, not a single repeated keyword phrase. Diverse anchor text reads more naturally to both Google and LLMs.
- 33. Orphan page auditAre any indexable pages disconnected from the internal link graph? Orphans almost never rank and are often forgotten content from old campaigns.
- 34. Hub-and-spoke topical clustersAre pillar pages connected to supporting content via bidirectional links? Topical authority compounds when the cluster structure is intentional.
Local SEO signals (35–38)
- 35. Google Business Profile completenessCategories accurate, services listed, photos current, hours correct, attributes complete, posts cadence active. GBP completeness directly affects Map Pack visibility.
- 36. NAP consistency across top 30 Canadian citation sourcesName, address, phone identical across Yelp, Yellow Pages, BBB, vertical-specific directories. Inconsistencies dilute local authority signals.
- 37. Review acquisition cadence and response rateSteady review acquisition over time, response on every review (positive and negative), no review velocity anomalies that would trigger Google scrutiny.
- 38. Local landing page coverageCity, neighbourhood, and service-area pages with substantive unique content, not templated near-duplicates. Templated location pages are increasingly downweighted.
AI / AEO citation readiness (39–43)
- 39. Brand citation share across ChatGPT, Perplexity, Gemini, Google AI OverviewsTracked via dedicated AEO tooling on the brand's top 50 buyer-intent prompts. Baseline measurement is the prerequisite for any improvement work.
- 40. Quotable content patternsAre key pages structured with the patterns LLMs preferentially extract — clear Q&A formats, definition leads, numbered or bulleted summaries, attributable statistics with sources?
- 41. Sourced statistics with dates and attributionOriginal data, surveys, or first-party metrics with publication dates and attribution. LLMs preferentially cite dated, sourced, original statistics over unsourced claims.
- 42. Off-page authority for AI enginesMentions on the high-authority sources LLMs preferentially weight (Wikipedia, major publications, vertical authority sites, professional bodies). Different from classical link authority signals.
- 43. AEO-friendly schema graphEntity-rich schema linking organisation, authors, content, and topics into a cohesive graph the engines can extract reliably. Covered in detail in our <a href="/ai-seo" class="text-primary hover:underline">AI SEO services</a> framework.
Off-page & authority (44–47)
- 44. Backlink profile quality auditDistribution of referring domains, anchor text profile, link velocity, presence of any toxic patterns. Disavow only when there is genuine evidence of negative SEO; aggressive disavow is usually unhelpful.
- 45. Brand mentions (linked and unlinked)Volume and trajectory of brand mentions across the open web, particularly on authoritative sources. Unlinked mentions still affect entity recognition by LLMs.
- 46. Competitor link gap analysisWhere are credible competitors earning links the brand is not? Often surfaces editorial or partnership opportunities the brand has missed.
- 47. Branded search trajectoryAre branded searches growing, flat, or declining? Branded search volume is the cleanest single proxy for actual brand demand and correlates with everything else.
What we do with the results
Forty-seven checks produce a lot of findings. The output of an audit is not the finding list — it is the prioritised 90-day remediation roadmap a brand can actually execute against. We typically structure deliverables as: critical-now fixes (week 1–2), foundational improvements (month 1), program-level work (month 2–3), and ongoing-discipline items (continuous). Most brands cannot execute everything at once; the audit's value is in what the brand should do first, not in the comprehensive list.
If you would like the full 47-check audit run on your site, with a written report and a prioritised 90-day roadmap, request a free audit. Three business days for SMB sites, 7–10 business days for enterprise. Or call (437) 900-3626 to talk through your situation first.
Toronto SEO Audit — FAQ
The questions Toronto buyers ask most often about how we audit before quoting.
