Canonical Tags SEO: Fix Duplicate and Redirected Pages

I find the same fault on most WordPress sites I audit, and it stays invisible until traffic is compared page by page. Two addresses serve the same content, Google picks one the owner did not expect, and the page that has been polished for months is not the page that has been earning the rankings.

The phrase used for it, duplicate content, sends people looking for a penalty that does not exist. In 13 years I have never seen a site penalised for having duplicate URLs. What I have seen is signals split between near identical pages, crawl time spent on addresses nobody wanted, and reports that are far harder to read than they need to be.

This is for the owner or in-house marketer who has heard the term and does not know whether they need a canonical tag, a redirect or nothing at all. Canonical tags SEO comes down to three honest points: duplicates are normal, the tag is a hint rather than an instruction, and the damage comes from conflicting signals on your own site rather than from the duplicates themselves.

What a Canonical Tag Actually Does

The problem it was invented to solve

A canonical tag is a line of code in the head of a page that names the address you want treated as the representative version. It exists because the same content can be reachable at several addresses, and a search engine needs to settle on one to show, rank and report on. Google describes the consequence directly: the canonical page is crawled most regularly, and duplicates are crawled less frequently, to reduce the crawling load on sites.

What it does not do, and one myth to drop

It does not remove the other address from your site, it does not redirect a visitor, and it does not override a redirect. It also does not clear a penalty, because there is no penalty to clear. Google says so in one line: "Some duplicate content on a site is normal and it's not a violation of Google's spam policies." The explanation sits on Google's page about what URL canonicalisation is, and it is worth reading before you touch your template.

How Duplicate URLs Appear on a Normal Site

HTTP versus HTTPS, www versus non-www

The most common pattern is also the simplest. The secured and unsecured addresses, and the www and non www versions, all resolve to the same site. If one of them is the version you want, redirect the others to it and point your internal links at it. Anything else leaves Google to make the decision for you.

Tracking parameters, filters and sorting

Marketing adds parameters, shop filters create variations by the dozen, and sorting options multiply them again. Google groups these well when the patterns are consistent and poorly when they are not. A parameter that identifies a campaign should never be the version you link to from your own navigation.

Old URLs, staging copies and print versions

Last year's blog addresses, a staging copy left resolving on a subdomain, and print friendly versions all serve almost identical text. Each one is a small leak of signal unless something on the page tells Google which version counts.

Canonical Tag or Redirect: Choosing the Right Tool

When a redirect is the better signal

If the duplicate should not exist at all, redirect it. A permanent redirect is a strong statement, and Google states what it means: "The 301 and 308 status codes mean that a page has permanently moved to a new location." Two addresses for one thing, where one was never intended for humans, is a redirect case every time. Google's guidance on redirects and Google Search is worth reading before you write yours.

When a canonical tag is safer

Keep both addresses and use a canonical tag when each has a legitimate reason to exist. A filtered view that customers bookmark and share, a variant used in an email campaign, a page that has to stay live for another system. You are not removing anything, you are telling Google which version is primary. When the pattern runs across a large catalogue, this is the sort of work where working with an SEO consultant on the fix is cheaper than getting a template wrong twice.

Using both, and why order matters

You can redirect and canonicalise across the same site, and Google notes that these methods can stack and become more effective when combined. Order still matters where they conflict. A canonical pointing at an address that then redirects is a contradiction, and Google resolves it by ignoring the hint and choosing for itself.

Writing a Canonical Tag Correctly

Self referencing canonicals and absolute URLs

Put a canonical tag on every page, including the preferred version, pointing at itself. That way every page in a set makes the same statement and none is left silent. Write the full address with its protocol rather than a relative path, because different systems interpret a relative value differently.

Google sets the same standard for your internal links: "When linking within your site, link to the canonical URL rather than a duplicate URL." Your navigation is a signal in its own right, and it should agree with the tag rather than contradict it. Google's page on specifying a canonical URL covers the methods side by side.

The HTTP header option for PDFs and other files

An HTML page carries the instruction in its head. A PDF cannot, so it carries the same instruction in an HTTP header instead. If you serve documents that Google indexes, the header version is how you point at the copy you want shown.

One method per page, not three

Pick one signal per page and apply it consistently. Three methods nominating three different addresses are worse than no signal at all, because they tell Google that nobody involved has made a decision. That is exactly the condition that produces split signals, which is the real cost of duplicate content on a live site.

Why Google Sometimes Ignores Your Canonical

The signals Google weighs, including sitemaps

Google sets expectations in a single sentence: "That is, indicating a canonical preference is a hint, not a rule." Other signals carry weight alongside it, including whether the address appears in your sitemap, whether the page is served securely, and how the two addresses relate through redirects.

Mistakes that create conflicting hints

The conflicts are usually easy to find once you look for them. A canonical on a paginated page pointing at the first page of the set. A canonical written with the wrong protocol. A sitemap listing both versions. Internal links pointing at the address you asked Google not to use. Any one of those will move the decision away from the tag.

Checking Which URL Google Actually Chose

URL Inspection and the indexed version

Inspect the address you believe is canonical and read the indexed version Google reports back. If it differs, you have found the conflict and you know what to fix. Changing the tag before that check is guesswork. Google keeps its own record of the relationship between the two addresses: "When you redirect a URL, Google keeps track of both the redirect source (the old URL) and the redirect target (the new URL)." That is why reporting can show both addresses for a while after a redirect, and why a redirect is not instant amnesia.

Reading the Pages and indexing reports

The indexing reports group addresses by state, and two states matter here. Alternate page with proper canonical means Google accepted your hint. Duplicate without a user selected canonical means it chose something else. Read the second one carefully, because it tells you your tag is being overruled rather than ignored.

A Fix Order for a Site With Too Many Duplicates

Group the addresses by pattern first: protocol, host, trailing slash, parameters, filters, and old paths that should have been retired. Decide the preferred address for each pattern, in writing, before changing anything. Then apply one method per pattern at template level instead of editing pages by hand. Correct the internal links next, because they are the cheapest signal to fix and the one Google reads most often. Then resubmit the sitemap and wait a few weeks before judging the result.

If the patterns run through a catalogue, a booking system or thousands of filter addresses, work through the full technical checklist in order rather than jumping to the template. The mapping step is where most of these projects succeed or fail.

Traps: Canonical Chains, Looped Redirects and Noindex

Three traps cause most of the mess I find. A canonical chain, where page A nominates B and B nominates C, which dilutes the signal and leaves the choice uncertain. A redirect loop, where two addresses point at each other and neither resolves. And a noindex tag on a page that other pages name as their canonical target, which tells Google to drop the very address you nominated as the representative version.

All three are found by reading the head of the page and the status code of the response, which takes a few minutes per pattern once you know where to look.

How to Test the Fix and What to Monitor

Test by fetching the address and reading three things: the canonical tag in the head, the status code of the response, and the destination of any redirect. Then inspect the URL in Search Console and watch the indexed version change over the following weeks. Monitor by comparing reporting before and after, and by checking that the clicks and impressions move across to the address you chose rather than staying on the duplicate.

Duplicate patterns are a technical mapping problem, and a technical SEO audit that maps duplicate URLs is the fastest way to get the full picture on a site with thousands of addresses. If you want to see the shape of the work before you commit, how these fixes look in practice is a fair read.

Canonical Tags FAQ

Will duplicate content get my site penalised?

No. Google's own canonicalisation documentation states that some duplicate content on a site is normal and is not a violation of its spam policies. The real costs are wasted crawl time, split signals between near identical addresses and reporting that is harder to read, which is why consolidation is still worth doing.

What is the difference between a canonical tag and a redirect?

A redirect sends a visitor and a crawler to another address, and Google treats a permanent redirect as a strong signal that the target should be canonical. A canonical tag leaves both addresses live and asks Google to treat one as the representative version. If the duplicate should not exist at all, redirect it. If both addresses serve a genuine purpose, use a canonical tag.

Where should the canonical tag point?

On each duplicate page it should point at the one address you want shown in search. The best practice is to put a self referencing canonical on that preferred page too, so every page in the set states the same answer, and to use absolute addresses with their protocol rather than relative paths.

Can Google ignore my canonical tag?

Yes, and its documentation is explicit about it: indicating a preference is a hint, not a rule. Google weighs other signals, including whether the page is served over a secure connection, redirects, and whether the address appears in your sitemap. If Google keeps choosing a different address, look for conflicting signals rather than adding more tags.

My site has thousands of filter and parameter URLs. Do I canonicalise all of them?

Not one by one. Group them by pattern, decide the preferred address for each pattern, then apply one consistent method at template level. Google's canonical and sitemap guidance also notes that these methods stack, so a template level canonical plus consistent internal linking is worth more than thousands of hand edited tags.

How do I check what Google chose?

Use the URL Inspection tool in Search Console to see the indexed version of a specific address, and read the pages and indexing reports to spot which addresses Google is treating as duplicates. Confirm the live HTML and the response headers of the pages you expected to be canonical before you change anything.