Guides / Technical SEO
Noindex and canonical: a decision guide for three common SEO audit findings
Download the PDF guideAn audit flags a page as noindex. Another page points its canonical somewhere else. A duplicate URL returns a redirect. The right response depends on what each page is supposed to do.
Before changing a directive, answer three questions: Should this URL appear in search? Is its content substantially the same as another page? Which URL should receive visitors from search?
1. Establish which control you are examining
| Control | Primary question | Important limitation |
|---|---|---|
| robots.txt | May this crawler request the resource? | It is not a reliable instruction to remove a URL from Google's index |
| noindex | Should this content be excluded from search indexing? | Google must be able to fetch the resource and see the instruction |
| rel=canonical | Which version represents duplicate or very similar content? | It communicates a preference; Google's selected canonical must be checked separately |
| Permanent redirect | Should visitors and crawlers move to a replacement URL? | It changes the visitor's route, not just a search preference |
On a small screen, scroll the table horizontally.
Google distinguishes crawling control from indexing control. Blocking a URL in robots.txt can stop Google from seeing its noindex instruction, and a blocked URL can still appear without its content. For private information, indexing directives are not access control; authentication belongs in the access design.
Source: Google: robots.txt; Google: noindex
The following scenarios are illustrative decision exercises, not findings from Layerly customers.
2. Scenario A: a thank-you page has noindex
The URL confirms an enquiry and offers no independent search value. The owner does not intend it to be a search landing page.
Proposed decision: retain the intended exclusion. Confirm that the success page is not being used as the service page, and that the actual service URL remains eligible for search.
Do not remove noindex merely to clear an audit warning. Conversely, do not treat noindex as protection for customer details. Sensitive submission data should not be exposed on a publicly accessible thank-you page in the first place.
Acceptance check: the service page has the intended indexing signals; the confirmation page has the intended exclusion; no confidential data is publicly exposed.
3. Scenario B: a campaign URL canonicalises to the main service page
The campaign parameter changes attribution, while the page's substantive content remains the same. The owner wants one main service URL to represent it in search.
Proposed decision: assess whether the canonical relationship fits that intention. Check the destination, content similarity, internal links and sitemap for conflicting signals.
A canonical pointing elsewhere is not automatically a defect. Google describes redirects and canonical annotations as strong signals, with sitemap inclusion a weaker one. These signals should communicate a consistent preference. A report should identify any conflict it actually observed.
Source: Google: canonicalisation
Acceptance check: the declared destination is appropriate, crawlable and not contradicted by the sampled signals. Later, check Google's selected canonical in indexed data. The live URL test cannot predict that selection.
Source: Google Search Console: URL Inspection
4. Scenario C: a distinct service page canonicalises to the homepage
The page describes a separate service, has meaningful independent content and is intended to attract relevant searches. A shared template has assigned the homepage as its canonical.
Proposed decision: investigate the template and intended indexing role before editing. If the page should stand independently, align its canonical and supporting signals with that decision. Check for an inherited noindex instruction as a separate issue.
Do not redirect every excluded service page to the homepage. That can remove the specific destination visitors need. Do not use noindex as a substitute for consolidating duplicate versions; Google recommends canonical annotations for that purpose within a site.
Source: Google: canonicalisation
Acceptance check: the live page exposes the agreed signals, the template does not overwrite them and the sampled internal links reach the intended URL. Record Google's later indexed interpretation separately from the implementation check.
5. Split completion into two stages
A deployment can correct HTML today while Search Console still describes an earlier crawl. Google distinguishes the indexed report from a live test. Both are useful, but they answer different questions.
Source: Google Search Console: URL Inspection
Implementation complete: the approved change is deployed and independently checked on the live URL.
Search processing observed: a later crawl or indexed report reflects the intended change, or provides evidence for further diagnosis.
This split lets the delivery team complete a verifiable correction without presenting Google's processing as something it controls.
Keep this decision worksheet
Record these fields: URL · intended search role · unique or duplicate · observed directives · declared canonical · Google-selected canonical and crawl date · proposed action · implementation retest · later index review
A useful audit preserves intentional configuration and isolates accidental conflicts. Before accepting any recommendation to remove, consolidate or redirect a page, require the intended search role and the supporting evidence in the same record.