Answer
What is a canonical tag and why does it matter?
A canonical tag is a line of code telling search engines the genuine address of a page, so duplicate versions do not compete with each other. It matters because when a site is rebuilt or migrated, these tags frequently keep pointing at old URLs that now return 404 — which instructs search engines that the real version of every page does not exist, and they respond by removing the pages from the index.
- What it looks like
- <link rel="canonical" href="...">
- Its job
- name the real version of a page
- Common fault after a rebuild
- points at the old, now-dead URL
- Consequence
- pages dropped from search results
- Visible to visitors?
- no — the site looks perfect
- Typical recovery time
- 2-6 weeks after fixing
What the tag is for
The same page can often be reached at several addresses — with and without a trailing slash, with and without www, with tracking parameters attached. To a search engine these can look like separate pages competing with each other.
The canonical tag resolves that by naming one address as the genuine one. In normal operation it should point at the page's own live URL, which is called a self-referencing canonical, and it is what almost every page should have.
The fault that hides entire sites
The damage happens during rebuilds and migrations. A site is developed with one URL structure, then launched with another — pages that were built as files become clean paths, or a staging domain becomes the live one.
The canonical tags, hard-coded into each page, are frequently left pointing at the old addresses. Those addresses now return 404. Every page on the site is therefore telling search engines that its real version is a page that does not exist.
Nothing about this is visible to a visitor. The site looks finished and works perfectly. It simply stops appearing in search results, and because the decline is gradual it is usually attributed to competition or to the market.
We found this on fourteen of our own sixteen pages earlier this year, some months after launch. It is the reason we now check it on every site before anything else.
How to check yours in two minutes
Open any page on your site, view the page source, and search for the word canonical. You will find a line containing a URL.
Copy that URL into a new browser tab. If it loads the same page you were just on, that page is correct. If it 404s, or loads something different, you have found a serious fault. Check three or four pages, not just the homepage — homepages are frequently correct while every interior page is wrong.
Fixing it means updating each tag to the page's own live address, then requesting re-indexing in Google Search Console. Recovery typically takes two to six weeks.
Related questions we get
Is a missing canonical tag as bad as a wrong one?
No. Missing is untidy and can cause duplicate versions to compete. Pointing at a dead URL is actively destructive, because you are giving a clear instruction that the page does not exist.
Does WordPress handle this automatically?
Modern SEO plugins usually do. The problem arises when canonical tags are hard-coded into page content — common when a site is delivered as static HTML and pasted into a page builder, which is exactly how the fault reached our own site.
How do I know if this is why my traffic dropped?
In Google Search Console, check Pages for entries excluded by canonical, and look at whether indexed page count fell after a redesign. A drop that begins with a launch date rather than gradually is a strong signal.
Can I just remove canonical tags altogether?
Better than pointing them at dead pages, but not ideal. Self-referencing canonicals are the correct state and prevent a separate class of duplicate-content problem.
More answers
Want this checked on your own site?
Send the URL. The audit is free, it takes us about twenty minutes, and it comes back in two working days.