Content audits fail when teams mistake inventory for analysis, ignore indexation signals, or force every page into uniform scoring. Understanding what breaks audits — from flawed categorization to misread user intent — turns a mechanical checklist into strategic portfolio management.
The most pervasive content audit mistake is stopping at the spreadsheet. Teams export a site crawl, tag a few columns for word count and title length, then declare the audit complete. That is inventory, not diagnosis. A functional audit answers why pages exist, what job each one performs, and whether that job still aligns with business priorities and search demand. You need to append indexation status from Search Console, map each URL to its parent topic cluster, note whether it ranks for its intended keyword or an unintended one, and flag if it cannibalizes a higher-authority sibling. Without this layer, you are simply counting rows. The fix is to treat the URL list as raw material. Build columns for search intent category, funnel stage, backlink count from an export, organic sessions over the trailing quarter, and a yes/no field for whether the page has a clear conversion path. Only then can you spot patterns like bottom-funnel pages with no internal links or top-funnel content accidentally marked noindex.
Many audits apply a universal quality rubric: word count above threshold, readability grade, keyword density, image alt attributes. Pages that fail get marked for rewrite or deletion. This logic breaks when you mix content types. A product comparison targeting commercial intent legitimately needs fifteen hundred words and structured data. A local service page for one Ottawa neighbourhood may perform perfectly at four hundred words if it includes the GMB-linked NAP, a neighbourhood keyword in the H1, and a simple conversion form. Scoring both against the same template misidentifies which pages actually underperform. The correction is to segment your sheet by content type first — commercial product pages, informational blog posts, location pages, tool or calculator pages, support articles — then define type-specific success criteria. A support doc succeeds if it reduces ticket volume and ranks for a long-tail troubleshooting query. A pillar guide succeeds if it attracts backlinks and feeds traffic to child pages through internal links. Measure each group against its own outcome, not a generic checklist.
Google Analytics shows you pages that received sessions. It will not surface orphaned URLs with no internal links, pages blocked by robots but still in the index leaking duplicate signals, or noindexed pages that consume crawl budget without delivering value. Relying solely on Analytics data means the messiest parts of your site stay invisible. You need a full-site crawl from Screaming Frog, Sitebulb, or similar, cross-referenced with Search Console's coverage report and index status. Look for pages marked noindex that also appear in the sitemap — configuration errors that confuse crawlers. Identify URLs returning 200 status codes but excluded from Analytics because they never earned a visit; these often turn out to be pagination stubs, empty category archives, or test pages left live. Check for redirect chains longer than two hops and orphaned pages with strong backlink profiles that you accidentally de-linked during a redesign. The highest-value fixes frequently hide in this gap between what you measure and what actually exists.
Audit templates often flag any two pages ranking for the same keyword as cannibalization requiring consolidation. This oversimplifies. If you have an introductory guide to workplace safety training and a separate page selling your workplace safety course in Toronto, both may rank for variations of that phrase — and should. The guide targets informational intent at the top of the funnel; the service page targets transactional intent. Merging them destroys the funnel. Real cannibalization occurs when two pages target the same intent and split authority, preventing either from ranking strongly. The diagnostic is to compare search intent, not just keyword overlap. Pull the actual queries each page ranks for from Search Console. If both pages appear for the same transactional query and neither breaks page two, you have a problem. If one owns the informational query and the other owns the commercial variant, leave them separate and strengthen the internal link from guide to service page. Canadian sites with bilingual content face a related pitfall: flagging the French and English versions of the same service page as duplicates, when hreflang and proper canonical setup make both necessary.
When a page shows zero Analytics sessions over six months, the instinct is to delete or noindex it. Before you do, export its backlink profile from Ahrefs, Majestic, or your tool of choice, and check Search Console for historical impressions. A page with no recent traffic may still hold ten referring domains and have ranked well two years ago before a site migration broke its internal links. Deleting it wastes that authority. The better move is to restore internal links, update the content to match current search intent, and redirect only if the topic is truly obsolete. Similarly, review internal link counts before removal. A page with no external links but fifty internal links may serve as a hub passing authority to child pages; cutting it can collapse a section of your site architecture. For content you do retire, implement 301 redirects to the closest thematic match, not a generic homepage redirect. Map old product pages to current product categories, defunct blog posts to updated articles on the same topic, and obsolete location pages to a broader regional page. This preserves link equity and user experience when old URLs surface in backlinks or bookmarks.
Content audits typically assess what is written, not how it is delivered. Pages that rely on JavaScript frameworks to inject primary content may pass a human review but fail to pass meaningful text to Googlebot if server-side rendering or pre-rendering is missing. Similarly, pages with render-blocking resources, slow Time to First Byte, or oversized image files degrade crawl efficiency, meaning Google samples your site less frequently and updates rankings more slowly. During your audit, append Core Web Vitals data from PageSpeed Insights or CrUX, and flag pages with LCP above 2.5 seconds or CLS above 0.1. Cross-reference server logs to identify high-traffic pages that Googlebot crawls disproportionately often — usually pagination or filter URLs you should be consolidating or blocking. For Canadian sites serving bilingual content, confirm that language-switching does not trigger full-page JavaScript reloads that reset render budgets. The fix often involves moving critical text into the initial HTML payload, lazy-loading below-the-fold images, and using rel=alternate hreflang in the head rather than injecting it via script.
After tagging underperforming content, many audits stall because no one defines when to update versus when to cut. A clear decision framework prevents paralysis. Refresh content when the topic still attracts search volume, the page holds backlinks or strong internal link equity, and the core information remains relevant with factual updates. Retire and redirect when the topic is obsolete, search volume has collapsed, the page has no inbound links, and no reasonable update would restore value. For borderline cases, check whether the page could be consolidated into a more comprehensive guide. If you have five shallow posts on different types of commercial leases, merge them into one pillar page with jump links and redirect the old URLs. If you have a detailed guide on a niche compliance requirement unique to Ontario, keep it even if traffic is low — it may convert at a high rate for a small, high-intent audience. Document your criteria in the audit sheet: columns for keep/refresh/merge/delete, a consolidation target URL if applicable, and a one-sentence rationale. This turns subjective judgment into a repeatable process and gives stakeholders a clear basis for approvals.
Run a full-site crawl with Screaming Frog or Sitebulb, then export the list of discovered URLs. Cross-reference this against your XML sitemap and the URLs that appear in Google Analytics over the trailing ninety days. Any URL the crawler found but that has zero internal links and does not appear in the sitemap is orphaned. These pages can still be indexed if Google discovered them via an external backlink, but they receive no internal authority and often rank poorly or waste crawl budget.
Do not delete it outright. First, check whether you can expand the page with substantive information that matches current search intent. If the topic is too narrow to justify a standalone page, identify the most closely related pillar or category page, move any unique details into that destination, then set up a 301 redirect from the old URL. This preserves the link equity from those backlinks while eliminating the thin-content signal.
Yes. Bilingual sites need separate audit tabs for English and French URLs. Tag each with its lang attribute and canonical setup to confirm hreflang is implemented correctly. Assess whether translated pages match the search intent in that language market — Quebec queries often differ in commercial intent and local terminology. Also verify that internal links from the French nav and content sections point to French pages, not defaulting to English, which dilutes crawl signals and user experience.
A comprehensive audit makes sense annually for most sites, or after a major migration, CMS change, or algorithm update that shifts rankings significantly. Quarterly mini-audits targeting new content published since the last review or pages that dropped out of position one through ten help catch issues faster. For large portfolios above a thousand indexed pages, consider rolling audits where you assess one category or topic cluster each quarter rather than tackling everything at once.
Screaming Frog or Sitebulk for crawl-level issues like redirect chains, orphaned pages, and missing meta tags. Google Search Console for indexation status, coverage errors, and which queries actually trigger impressions. Ahrefs or Semrush for backlink counts and historical ranking data. PageSpeed Insights or Lighthouse for render and Core Web Vitals issues. Combine exports from all of these into your master audit sheet so you have technical, link, and performance context alongside content quality judgments.
Pull Search Console data for a full twelve-month period, not just the most recent quarter. Look for impression spikes at specific times of year. For example, a guide on RRSP contribution deadlines in Canada will see traffic in February and March but almost none in summer. If you audit in July using only ninety-day data, it looks dead. Checking the annual pattern reveals its seasonal value. Tag these pages to refresh before their next peak season rather than marking them for deletion.