A technical audit should explain what prevents useful pages from being discovered, understood or used—and what to fix first. A crawler export is an input to that judgment. It is not a prioritized action plan by itself.
Begin by defining the site’s important page types and outcomes. A publication, a store and a documentation library have different routes and failure costs. Record the audit scope, crawl settings and observation date so another person can reproduce your findings.
Build a representative inventory
Collect published URLs, sitemap entries and URLs found by a crawl. If available, add landing pages from search reporting and verified crawler requests from server logs. Each source has blind spots. A crawl cannot reveal every orphan page, and a sitemap can contain outdated URLs.
Group pages by template: homepage, categories, articles, products, filters and other relevant types. Sample each group manually. Template-level problems usually deserve attention before polishing isolated pages. The architecture guide describes how to compare a publishing inventory with the reachable site.
Check access and HTTP responses
Important public pages should return the expected content and a suitable status. Inspect server errors, redirect loops and broken internal destinations. Confirm that moved pages redirect to an equivalent resource. A removed page without a replacement should not be redirected indiscriminately to the homepage.
Check robots rules and login requirements, then inspect the rendered page. Is meaningful content present? Are important links ordinary discoverable anchors? Do mobile visitors receive the same essential information? Separate crawl discovery and retrieval from later indexing questions.
Review indexing and canonical signals
Look for accidental noindex directives on intended landing pages. Compare canonical references, internal links and sitemap URLs. Conflicting preferences make diagnosis harder. A canonical points to a preferred representative; it is not a substitute for redirecting a permanently moved page, and search engines can choose differently.
Inspect duplicates caused by parameters, alternate protocols, print views or repeated content. Decide which versions have a legitimate purpose before applying a blanket rule. Google’s canonicalization guidance explains the roles of redirects, canonical annotations and sitemaps.
Audit internal links and site structure
Find important pages with weak or missing inbound routes, links to redirects and navigation that omits useful sections. Test pagination and archives rather than assuming that the first category page exposes every article. Review anchor text where similar destinations are hard to distinguish.
Use a focused internal linking review to turn those findings into template fixes and editorial updates. An orphaned section may need a hub and navigation change, not a larger sitemap.
Inspect performance and mobile usability
Check loading, visual stability and responsiveness on representative templates. Look for oversized images, unnecessary scripts, delayed main content and interface elements that move while a visitor interacts. Combine field evidence, when available, with reproducible lab tests. One lab run is a diagnosis aid rather than a universal experience score.
At narrow widths, test the actual menu, long titles, tables, search controls and footer. Confirm visible keyboard focus, readable text and usable touch targets. A page fitting inside the viewport does not prove the interaction works.
Validate structured data against visible content
Structured data should describe information actually present on the page. Use an appropriate type, validate the markup and inspect warnings in context. Do not add invented ratings, people or organizations to satisfy a tool. Valid markup does not guarantee a special search appearance; see Google’s structured data introduction.
Prioritize findings instead of counting them
| Priority | Example | Next action |
|---|---|---|
| Urgent access failure | Important templates return server errors or accidental noindex | Confirm scope, repair and retest immediately |
| High-impact structural issue | A useful section is disconnected from navigation | Fix the shared route and verify affected pages |
| Targeted improvement | Some internal links use old redirecting URLs | Update references in a controlled batch |
| Low-confidence observation | A tool flags a pattern without a demonstrated problem | Investigate before assigning development work |
For each finding, write the affected URLs, evidence, likely consequence, proposed fix, owner and acceptance test. Include confidence and effort. A small issue on a critical conversion path may matter more than a large count of harmless warnings.
Verify the fix and its outcome
After deployment, repeat the original check and sample adjacent templates for regressions. Record the release date, then monitor relevant search and usage signals over a suitable period. A corrected response can be verified immediately; search-engine processing and traffic outcomes take separate observation.
Audit deliverable: a short, evidence-backed repair queue with clear acceptance tests. Connect that queue to meaningful SEO measurement so the audit improves the site rather than simply reducing a tool’s warning count.