What It Actually Means
That happens more than people expect, usually without anyone intending it. A page might be reachable with and without www, over http and https, with and without a trailing slash, with tracking parameters appended, with filters or sort orders in the address, or as both a category page and a direct page.
To a search engine those can look like separate pages carrying identical content. The canonical tag resolves it: treat this one as the original, credit it with any signals, ignore the rest.
Why It Matters Commercially
Two reasons.
Signals get split. If three addresses show the same page and links point at different ones, the value is divided across them instead of accumulating on one. Neither version is as strong as the single page would have been.
Attention gets wasted. Every duplicate address consumes crawl budget, which matters a great deal on a site that is already being rationed.
Where It Goes Wrong
This is a small tag capable of causing large and quiet damage, which is why it is worth checking rather than assuming.
Every page pointing at the homepage. A configuration error that appears occasionally, usually after a migration or a plugin change. The effect is that the entire site tells search engines only the homepage is real, and everything else disappears from results over the following weeks. It is silent, and the cause is not obvious from the symptom.
Pointing at the wrong version. Canonical says https, links and sitemap say http. Mixed signals produce unpredictable outcomes.
Canonical to a page that redirects. The chain has to be followed and the instruction is ambiguous.
Canonical to a noindexed page. Contradictory: this is the real version, and do not index it.
Missing entirely on a site with parameters. Tracking parameters on campaign links can generate a large number of duplicate addresses.
The Simple Rule
Every page should carry a canonical tag naming itself, using the exact address you want indexed, absolute rather than relative, matching what the sitemap says and what internal links point to.
Consistency across those three is the whole discipline. Most canonical problems are not conceptual, they are one of the three disagreeing with the others.
What To Ask
- Does every page canonical to itself, with the correct address? Check a handful across different templates.
- Do the canonical, the sitemap and our internal links agree? Disagreement is the usual root cause.
- Are tracking parameters generating duplicate addresses? Frequently yes after any campaign.
- Have indexed pages dropped off unexpectedly? Worth checking canonicals first. It is a common cause and an easy one to miss.
Where It Does Not Help
A canonical tag is a strong hint rather than an instruction, and search engines can disregard it if the pages differ meaningfully.
It also does not solve genuine duplication, where two pages really do cover the same intent with different words. That is a content problem, and the answer is consolidating them into one better page rather than labelling one as secondary. Using canonicals to manage self-competition that should have been resolved editorially leaves both pages weaker than a single page would have been.
More terms are in the glossary.
