Canonical Tags and Duplicate Content, Explained Simply

By Neha·Published October 5, 2026

Here is something that surprises most website owners: your site almost certainly has more pages than you think. Not because anybody wrote extra ones. Websites make copies of their own pages by accident, all the time, and each copy has its own address.

Google then has to choose which address to show. Sometimes it picks the one you would have picked. Sometimes it picks a strange one with tracking codes hanging off the end, or it spreads the credit across several copies so that none of them ranks as well as one clean page would.

That is the duplicate content problem, and the canonical tag is the main tool for solving it. Both terms sound far more frightening than they are.

First, the myth: there is no duplicate content penalty

A lot of owners have been told that Google punishes websites for duplicate content. For ordinary sites this is not true. Google has said for years that accidental duplication inside your own website does not earn a penalty.

What actually happens is duller and still costs you:

• Google picks one version to show and hides the rest. It may not pick your favourite.
• Links pointing at different copies are counted separately instead of being added together.
• On a large site, Google spends its visits reading copies instead of finding your new pages.

The real penalties are reserved for sites that deliberately copy other people's content at scale. A shop whose product page exists at four addresses is not in that category. It just has housekeeping to do.

Five ways your website copies itself

1. The front-door problem. These four addresses often show the same home page: http://yoursite.in, https://yoursite.in, http://www.yoursite.in and https://www.yoursite.in. To a person they are one website. To a machine they are four. Only one should load, and the other three should redirect to it.

2. Slashes and capital letters. /services and /services/ can be two addresses. On some servers so are /Services and /services.

3. Tracking tags. When you share a link with campaign tags on the end, like ?utm_source=whatsapp, that is technically a new address showing the same page. The same goes for the codes Facebook and Google Ads add automatically. Tracking is worth doing, and our guide to tracking marketing with UTM links shows how, but the tagged versions must not compete with the clean one.

4. Filters and sorting on shops. An online store is the worst offender. One category of kurtas can be sorted by price, filtered by size, filtered by colour and split over several pages. Each combination produces a new address with nearly the same products. A shop with 200 products can easily show Google 5,000 addresses.

5. The same product in two categories. A blue cotton kurta listed under both "Kurtas" and "Summer Collection" may have two addresses, one through each category.

There are quieter ones too: printer-friendly versions, a test copy of the site left open on a subdomain, and near-identical location pages where only the city name changes.

What is a canonical tag?

A canonical tag is one line in the hidden top section of a page that says: "the official address of this page is this one." Canonical simply means official or preferred. It looks like this:

<link rel="canonical" href="https://yoursite.in/kurtas/blue-cotton-kurta" />

Put that same line on every copy, including the official page itself, and Google understands that all of them are one page and that the credit belongs to the address named in the tag.

An everyday comparison: a shopkeeper with one shop and three signboards on three different lanes. All three boards carry the same line, "Main entrance on M.G. Road". Customers arriving from any lane end up at the same door.

Two details are worth knowing. First, a page should normally name itself as canonical. This is called a self-referencing canonical, and it protects the page from tracking-tag copies. Second, the tag is a strong hint, not a command. If your canonical says one thing while your sitemap and internal links say another, Google may overrule you.

Canonical tag or redirect: which one do I need?

They solve different problems, and mixing them up is common.

A redirect moves people. A canonical tag leaves people where they are and only tells search engines which copy counts. If you need to write redirect rules, the redirect generator turns a list of old and new addresses into rules for your server and catches mistakes. The background is in our guide to broken links and redirects.

How to check your own site

You can do a useful check in fifteen minutes.

1. Test the front door. Type each of the four versions of your home address into the browser. All should end up at the same one, with https. If two of them load separately, ask your developer or host to add a redirect.
2. Check the canonical on your main pages. Paste a page into the canonical tag checker. It shows which page the address claims is the real one and flags conflicts, such as a canonical that points to an error page or to the wrong page entirely.
3. Look for repeated titles. Duplicate pages nearly always share a title. The duplicate title and description finder reads up to 25 pages and groups the ones that match. When the pages are genuinely different and only the titles repeat, our guide to duplicate titles and meta descriptions shows how to give each one its own.
4. Ask Google. In Search Console, open the Pages report and look for two lines: "Duplicate without user-selected canonical" and "Duplicate, Google chose different canonical than user". The first means Google found copies and you gave no guidance. The second means you gave guidance and Google disagreed. Our Search Console guide shows where the report lives.
5. Compare two suspicious pages. If you suspect two pages are too alike, the page comparison tool puts them side by side.

Mistakes that cause real damage

• Every page pointing to the home page. A template error that tells Google your whole site is one page. Service pages then disappear from results. This is the most harmful canonical mistake and it is surprisingly common after a redesign.
• Canonical pointing to a page that redirects or does not exist. The signal leads nowhere, so Google ignores it.
• Canonical plus noindex on the same page. One says "count this as that other page", the other says "do not show this". Mixed messages get unpredictable results.
• Using http in the tag on an https site. A leftover from before the site got its security certificate.
• Using canonicals to hide near-identical city pages. If you have fifty pages that differ only by the city name, the tag is not the answer. Fewer, better pages are. The same applies to language versions, which use hreflang tags, a different mechanism that the hreflang checker can test.

Fixing duplicates is one part of keeping a site tidy. It sits naturally in a regular SEO audit alongside your on-page SEO checklist.

Frequently asked questions

Will Google penalise my site for duplicate content?

Not for accidental duplicates within your own site. Google simply chooses one version to show. The cost is weaker rankings and wasted crawling, not a penalty. Deliberately copying content from other websites is a different matter and can lead to action.

Does every page need a canonical tag?

It is good practice for every indexable page to carry one that points to itself. Most modern platforms add it automatically. It protects the page from copies created by tracking tags and other additions to the address.

I use the same product description as the manufacturer. Is that a problem?

It is not penalised, but it gives Google no reason to prefer your page over the hundred other shops using the same words. Rewrite at least your best sellers in your own words, and add details only you know, such as sizing advice or delivery times.

How long does Google take to respect a new canonical tag?

It has to revisit both the copy and the official page, which can take from a few days to several weeks. In Search Console, the URL Inspection tool shows which address Google currently treats as canonical for any page.