Compare the URL you want indexed with Google’s selected canonical and the page’s actual signals. Check redirects, canonical markup, internal links and sitemap entries together. Use the indexed result and live test for their different purposes, then allow for recrawling.
A canonical tag is a preference. It is not a command.
When Google chooses another URL, inspect the disagreement before changing the template.
1. Identify the intended page
Copy the exact URL, including hostname and parameters. Write why it should represent this content: the main service, product or article page. A generated URL is not automatically a distinct page worth indexing.
Use the four pillars of SEO to place this technical check within the wider plan.
2. Separate indexed and live evidence
Inspect the URL in Search Console. Record status, last crawl, declared canonical and Google-selected canonical where available.
Google's URL Inspection guide distinguishes the stored indexed version from the live test. A successful live test checks current access and potential indexability. It does not prove indexing or predict canonical selection. If the crawl predates a release, record that timing.
3. Compare the selected URL
Open and inspect Google's selected address too. Compare substantive content, response codes, redirect destinations and canonical declarations. Look for parameter duplicates, another hostname or a legacy route.
Document whether the pages serve the same purpose. Do not merge distinct useful pages simply to make the export shorter.

4. Align the site's signals
Google's canonical guidance describes redirects and canonical annotations as strong signals; sitemap inclusion is weaker. Choose the appropriate treatment for the actual duplicate or retired page.
Point relevant internal links and sitemap entries to the intended address. Use crawlable links with real destinations. Check rendered output after a website change, not just the template source.
5. Check access, then verify the repair
Inspect HTTP responses, robots rules and indexing directives separately. Google's robots.txt guide explains that crawling controls do not reliably prevent indexing. Its robots meta guidance covers directives the crawler must be able to read. Private pages need actual access protection.
Fix the conflict you found and retest the published page. Request indexing when appropriate, then inspect the next crawl. A completed release and Google's subsequent selection are separate observations.
Inspection record: intended URL; selected URL; crawl date; response; canonical; access rules; conflicting signals; repair; next check.
Use the Search Console operator's guide for recurring review. For technical SEO, identify the affected URLs and the conflicting signals before changing a shared template.
Fill the inspection record before asking for a fix
Use the paired-URL example in our duplicate-content guide and download its canonical inspection sheet. Here is how to populate the Google-specific fields.
Paste the exact affected URL into Search Console's inspection bar and run the inspection. Expand the indexing details and copy the user-declared and Google-selected canonical values where available, plus the last-crawl date. Repeat for the preferred URL. Missing information stays “not available”; it does not mean that Google agrees with your preference.
Illustrative result: the parameter URL declares itself, Google selects the clean URL, and both show the same product. This may already be a sensible Google selection. The repair is the site's conflicting declaration and internal link—not an attempt to force Google to index both copies. If instead Google selects an unrelated collection, first compare the actual product content and inspect whether the intended page redirects or has an access problem.
Keep the two checks separate
| Observation | What it establishes | What to do next |
|---|---|---|
| Live page returns 200 and declares the intended canonical | The current response contains that signal | Check internal links, sitemap and the competing URL |
| Live test succeeds, but the indexed crawl predates the repair | Current access may be fixed; the stored result is older | Keep the indexed outcome pending and inspect a later crawl |
| Google selects the clean URL while a duplicate declares itself | The site's declarations disagree with Google's selection | Compare the content and align the duplicate's signals |
| Selected URL serves a different purpose | This may involve more than an ordinary duplicate | Investigate content, redirects and access before consolidating |
Use the sheet's live and indexed fields separately. “Not available” is a recorded limit, not a pass.
After the developer changes the conflicting signal, open the published URL and repeat the live check. Record, for example, “200; clean product canonical; category link corrected” in the live-result column. Leave the indexed-result column pending until the inspected crawl reflects the change. A screenshot of the deployment succeeding cannot fill that column.
Sources
- URL Inspection guidesupport.google.com
- canonical guidancedevelopers.google.com
- crawlable linksdevelopers.google.com
- robots.txt guidedevelopers.google.com
- robots meta guidancedevelopers.google.com
