Should you use a canonical URL or redirect a duplicate page?

One product can have several web addresses: a clean product link, a campaign link with tracking parameters, and an older address left behind after a redesign. The business question is not simply whether those URLs look untidy. It is whether each address still serves a purpose, and which one should represent the content in search.

1. What does a canonical URL actually decide?

A canonical URL is the address selected to represent a group of pages with duplicate or very similar main content. Google makes its own selection, called the Google-selected canonical. Your declared canonical tells Google which address you prefer, but it does not compel Google to choose it. Ordinary duplication is not automatically a spam-policy violation. [1]

Duplication can come from HTTP and HTTPS versions, alternate device versions, sorting options or campaign parameters. However, different addresses can also contain genuinely different material. A translated page is not automatically a duplicate of its English equivalent merely because both describe the same service. [1]

For a business website, canonical URL SEO is largely about consistency. Choose the address you want customers to find and share, then align the site's signals around it. This can help consolidate signals associated with duplicate pages, but it is not a shortcut to guaranteed rankings or enquiries. [2]

2. Choose a canonical, a redirect or a separate page

Start with the visitor's needs. Should someone opening the original address stay there, move to another address, or see content that deserves its own page? That question usually makes the technical choice clearer.

Google treats redirects and canonical annotations as strong canonicalisation signals. Permanent server-side redirects are appropriate for permanent moves; both 301 and 308 communicate that change. Temporary 302 and 307 redirects do not provide the same signal that the destination should become canonical. [2][3]

A canonical annotation does not redirect a visitor. That makes it useful when an alternate version must remain accessible, but unsuitable as the only fix for an obsolete address. Conversely, redirecting every variation can remove useful visitor behaviour. Decide what should happen before applying a site-wide rule.

3. Make canonical tags, links and sitemaps agree

For an HTML page, a canonical link annotation belongs in the page's head section and should use an absolute address. For example, a duplicate product page could contain <link rel="canonical" href="https://example.com/products/brass-diya">. The preferred page can use the same annotation to point to itself, known as a self-referencing canonical. [2]

Internal links should normally point directly to the preferred address. Your sitemap should also list the URLs you want appearing in search, using fully qualified addresses. Sitemap inclusion is a weaker canonicalisation signal than a redirect or canonical annotation, so it supports the decision rather than replacing those mechanisms. [2][5]

• The preferred page is accessible and contains the intended content.
• Duplicate pages declare the same preferred address rather than competing alternatives.
• Navigation, product links and related-content links use the preferred address.
• The sitemap lists preferred URLs rather than campaign duplicates or retired addresses.
• A changed URL has its redirect, canonical annotation and sitemap entry reviewed together.

4. Treat pagination and filters according to their content

A question mark in an address does not make the page disposable. A category URL ending in ?page=2 may show products missing from page one. Google recommends a separate URL and canonical for each page in a paginated sequence, rather than pointing all pages to the first. [4]

Connect those pages with ordinary crawlable links using an anchor element and an href address. A 'Load more' button alone is not enough: Google's crawlers generally do not click buttons or trigger JavaScript that requires user interaction. Google also no longer uses rel="next" and rel="prev" to identify pagination relationships. [4]

Filters and sorting need a separate assessment. Does the variation simply rearrange the same material, create a useful distinct selection, or produce a page you do not want indexed? Google discusses noindex for excluding unwanted filter or sort variations, but exclusion is a different decision from choosing a canonical. Do not remove every query parameter without checking what it changes. [2][4]

5. Hypothetical worked example: a homeware catalogue

This illustrative example is not a Softglimmer client case study. A Jaipur homeware manufacturer sells brass diyas and shares product links through WhatsApp campaigns. Its team finds five URL types and wants a rule that preserves useful pages without promoting duplicate addresses.

• /products/brass-diya: keep as the preferred product address and give it a self-referencing canonical.
• /products/brass-diya?utm_source=whatsapp: retain the campaign URL and tracking, but declare the clean product address canonical because the main content is identical.
• /products/traditional-diya: permanently redirect this retired address to /products/brass-diya because it represents the same product.
• /collections/brass-diyas?page=2: keep its own canonical because it lists additional products, and link it from the paginated category.
• /products/brass-diya-large: assess independently because different dimensions and ordering information may justify a separate product page.

These are practical applications of Google's duplicate-content, redirect and pagination guidance, not a promise that every product variation deserves its own search listing. The team should update internal links and the sitemap to use preferred product addresses, then inspect the clean product page and a campaign duplicate to check Google's selection. [1][2][3][4][5][6]

6. Implement a small, testable set of rules

1. Collect examples of clean URLs, campaign links, retired addresses, filters and pagination. Start with important product or service pages.
2. Compare the main content and visitor purpose. Record whether each variation needs a canonical, redirect or independent page.
3. Choose the preferred address for each duplicate group and brief whoever maintains the website.
4. Apply the rules and update internal links and sitemap entries in the same release.
5. Check the actual page output and redirect behaviour, not just a settings screen.
6. Use Google Search Console's URL Inspection to compare your declared canonical with Google's selected canonical. Request re-indexing for important corrected pages where appropriate, then check again after Google has reprocessed them. [6]

Avoid using robots.txt to choose a canonical, or adding noindex merely to force Google towards another duplicate. Do not declare conflicting canonical destinations or send unrelated pages to one preferred URL. If pages intended to remain separate keep being grouped together, investigate both technical errors and whether their main content is meaningfully different. [2][6]

Your next step is a one-page URL decision sheet covering your most important pages and their known variations. For a wider migration, use Softglimmer's website redesign SEO checklist at /insights/website-redesign-seo-checklist. If the rules need technical review, explore Website Performance & Technical SEO at /services/seo-performance or discuss the scope at /contact.

7. Frequently asked questions

This research-assisted guide uses the primary sources linked above. Examples are illustrative unless explicitly identified otherwise. Check original documentation when implementing changes.