Technical SEO · proof-based audit
SEO Audit Checklist 2026: Crawl, Indexing, Content, Priorities and QA
A useful SEO audit is not a collection of 300 red warnings from a crawler. It is a reproducible list of affected URLs, a plausible impact, evidence, owners and a verification method. This guide turns those ingredients into a delivery plan without pretending that one tool's health score predicts traffic.
01 / 6
Start with an inventory tied to outcomes
Group URLs by template and user goal: product purchase, enquiry, educational reading or subscription. Not every parameter URL requires indexing, and an error on a high-value product route differs from one on an obsolete filter page.
Merge sitemap URLs, internal HTML links, Search Console landing pages and server logs where available. Compare those sets to locate orphan pages, unexpected status codes and canonical pages that visitors can no longer reach through navigation.
Acceptance criteria
- Each indexable template has a business goal.
- Important orphan and non-sitemap URLs are identified.
02 / 6
Read real server responses
Crawl actual production URLs and record final status, redirects, HTTP headers, meta robots, canonical, hreflang, title, H1 and substantial text. A browser screenshot after hydration doesn't prove what an HTTP client receives.
Non-existent content should produce a genuine 404/410, moved URLs a specific 301, and current content a 200 with useful HTML. Redirecting every removed product to the homepage or serving a blank SPA shell with 200 is not a content solution.
Acceptance criteria
- Missing pages return correct 404/410.
- Key landing pages serve useful initial HTML.
03 / 6
Separate raw HTML from rendered DOM
Fetch a representative URL per template without executing scripts; then compare it with the DOM after JavaScript. Look for missing body text, changing canonical/robots signals, links behind interactions and broken data fetches.
Google can render JavaScript; that does not make asynchronous content failures harmless. Record the exact failing URL and API state before prescribing a costly migration to full SSR.
Acceptance criteria
- Initial HTML and hydrated DOM differences documented.
- Root cause attached to each rendering bug.
04 / 6
Consolidate URL signals and repair internal links
Check that redirects, canonical, sitemap, hreflang and navigation consistently point to the intended URL. A rel=canonical tag is a signal, not a command Google must follow.
Build a path from the homepage and topic hubs to priority pages using crawlable HTML anchors. Avoid orphan articles, endless faceted URLs and JavaScript-only buttons that never expose href links to users or search engines.
Acceptance criteria
- Canonical, internal links and sitemaps agree.
- Revenue and editorial pages receive contextual HTML links.
05 / 6
Use GSC, logs and editorial review as independent inputs
Search Console shows impressions, clicks and Google's known indexing states; access logs show actual HTTP fetches. A crawler exposes another perspective. Keep the sources separate rather than assigning a single made-up Google health number.
Inspect article usefulness as a distinct workstream: does the page answer a real intent with verified examples, sources, limitations and an independently useful outcome? Rewording a generic paragraph ten times is not a content strategy.
Acceptance criteria
- Every finding includes a dated source of evidence.
- Human usefulness is assessed separately from crawl validity.
06 / 6
Turn findings into tests and track impact
Prioritise broken indexability and critical transactions before low-impact styling recommendations. Define problem, affected template, current HTTP evidence, intended result, owner, due date and a smoke test for each issue.
After deployment verify the fix immediately on production, then wait for Google's own reporting and search demand data before claiming growth. An improved HTML response is a verified technical achievement, not proof of additional Google traffic.
Acceptance criteria
- Each change has a named owner and pass/fail test.
- Technical improvement and ranking outcomes are recorded separately.
Need to inspect title tags and meta descriptions before measuring CTR? Use the browser-local SEO Title & Description Checker with an illustrative SERP preview.
Are crawl restrictions part of the problem? Test Allow/Disallow rules for your URLs in the robots.txt checker.
Copy and adapt
Ready-to-use URL-level audit worksheet
Paste into Sheets and replace all example values after running your own checks. This CSV is not a crawl report for a real domain.
url,template,goal,status,initial_html,canonical,inlinks,gsc_state,evidence,impact,issue,owner,acceptance,status_work
https://example.com/service,service,lead,200,visible,self,9,unknown,curl,high,verify real lead,developer,lead exists in backend,open
https://example.com/guide,article,reading,200,missing,self,1,crawled_not_indexed,HTTP+GSC,medium,investigate JS rendering,developer,source HTML contains full article,openVerify completion
0/12Mark only steps you verified on your own site. These checkboxes reset on reload and do not certify the result.
Common failures and fixes
A crawler reports 100% health yet organic visits stall
Cause: Tool checks don't measure competitiveness, demand or the uniqueness of answers.
Fix: Analyse query cohorts, actual SERPs and the value created for readers.
Unknown URLs return a 200 page
Cause: SPA fallback intercepts every route and serves a shell.
Fix: Implement actual 404 at the HTTP response level.
Fix is deployed but Search Console hasn't updated
Cause: Google reports its own latest crawling, not your release timestamp.
Fix: Save proof of the HTTP fix and monitor inspection plus performance separately.
Questions people ask
What tool should I use for an SEO audit?
Use a combination of crawling, GSC, logs and manual testing. No single tool establishes indexation, value to readers and business results at once.
What should a small site audit first?
Indexable URL inventory, canonical tags, HTTP status, meaningful initial HTML, useful links and Google Page Indexing.
Can an audit guarantee traffic growth?
No. A technical fix can be proven by tests; search ranking and demand outcomes need separate observation.
Method, limits and source evidence
Reviewed October 15, 2026. The workflow combines URL inventories, HTTP, DOM, Search Console/logs, editorial review and acceptance tests.
This is a reusable audit methodology, not a claim that any particular client's site was crawled or tested.
Read the corresponding book chapters 5–9, 32 and 39 for deeper author explanations.
Found a factual error? Contact the author.
Read the original book chapters
These subjects are explored in the original SEO 2026 book. Full chapters are freely readable with citations, original tables and figures.