Duplicate and near-duplicate URLs are easy to create without deliberately publishing duplicate content. Tracking parameters, filtered pages, print versions, product variations and different linking conventions can all leave several addresses showing substantially the same material. A canonical tag helps a site indicate which URL should be treated as the preferred version. For a small business managing SEO, the useful part is not merely knowing the HTML syntax; it is deciding when canonicalisation reflects the site accurately and when it hides a structural problem that should be fixed another way.
Decide which URL deserves to be canonical
The preferred URL should normally be the stable address you want people to visit, internal links to use and search engines to consider for indexing. If several URLs display the same article because campaign parameters have been added, the clean article URL is an obvious candidate. Other cases need more judgement.
Two pages that happen to mention similar services are not automatically duplicates. If each serves a genuinely different search need, forcing one to canonicalise to the other may conflict with the purpose of keeping both pages. Before adding a tag, establish whether the URLs are alternative versions of essentially the same content or distinct pages that need stronger differentiation.
Add the canonical element in the page head
The canonical instruction is placed in the HTML head using a link element whose relationship is canonical and whose href identifies the preferred URL. In practical terms, the page template or content management system often generates this automatically. That is usually safer than manually editing every page, provided the generated value is correct.
Use the complete preferred URL and keep its format consistent with the live site. Differences involving protocol, hostname or trailing-slash conventions can create needless ambiguity. If a platform offers an SEO field for canonical URLs, check what it actually outputs in the rendered source rather than assuming the label guarantees the intended result.
Self-referencing canonicals can make the preference explicit
A unique page can point its canonical tag to its own preferred URL. This self-referencing approach gives a consistent signal about which address represents the content, particularly when the same page can be reached with additional parameters.
For a small site, implementing canonicals consistently through the template also reduces the risk that only pages considered problematic receive them. The important requirement is accuracy. A template that automatically generates the wrong hostname or an outdated path can reproduce the same mistake across the entire website.
Do not confuse canonical tags with redirects
A canonical tag does not send the visitor anywhere. If somebody opens a non-canonical URL, that URL can still load while the markup indicates a preferred version for search purposes. A redirect behaves differently by moving the browser to another address.
If an old page has permanently moved and there is no reason for visitors to keep accessing the old URL, a permanent redirect may be the clearer solution. Canonicalisation is more useful where alternative URLs need to remain accessible or are created by the way the site functions. Choosing between the two starts with the desired visitor behaviour, not with whichever SEO setting is easiest to find.
Keep internal signals aligned
If the canonical says URL A is preferred while navigation repeatedly links to URL B, the site is sending mixed messages. Internal links, sitemap entries and other controllable references should generally use the preferred address. Consistency also makes analytics and maintenance easier because traffic is less likely to be fragmented across avoidable URL variants.
For ecommerce and filter-heavy sites, this deserves planning before changing settings at scale. Some filtered combinations may provide useful standalone landing pages, while others are merely alternative views. Treating every parameterised URL identically can remove useful distinctions or leave unnecessary combinations competing for attention.
Test the rendered page, not only the CMS field
After adding canonical tags, inspect the live HTML and confirm that each page points where expected. Sample different page types rather than checking only the homepage. Articles, services, products, categories and paginated or filtered sections may use different templates and therefore behave differently.
Look for obvious defects such as multiple canonical elements, canonicals pointing to redirected or missing pages, staging domains appearing in production, or large groups of pages all naming one unrelated destination. A crawl of the site can make these patterns easier to spot than manual checks alone.
Treat canonicalisation as part of URL governance
Canonical tags work best when the underlying website has a deliberate URL policy. Decide how preferred addresses are formed, how parameters are handled, how internal links are generated and what happens when content genuinely moves. Documenting these decisions prevents a later redesign or plugin change from quietly reversing them.
For owners doing their own SEO, the goal is not to add a canonical tag because an audit tool expects to see one. It is to make the preferred version of a page clear and keep the rest of the site consistent with that choice. The tag is one signal within that wider discipline.