Should you really choose between www and non-www for your domain?

Why does Google impose this choice between www and non-www?

For search engines, example.com and www.example.com are considered two distinct entities. Each has its own backlink profile, crawl history, and ranking signals. Without explicit direction, Google may index both versions simultaneously, creating duplicate content that dilutes your performance.

The 301 redirect serves as an unambiguous signal: it indicates which version should inherit all accumulated SEO properties. Google then transfers PageRank, backlinks, domain age, and user signals to the canonical URL. This transfer is never instantaneous but occurs gradually during recrawl phases.

What is the difference between a 301 redirect and a canonical tag?

The 301 redirect operates server-side: it sends a permanent HTTP code that forces the browser and bots to the target URL. It is the most reliable method to consolidate signals. The canonical tag, on the other hand, remains a mere suggestion that Google may ignore if conflicting signals are present.

In the case of www/non-www, the canonical tag alone is insufficient. Users manually entering the non-redirected URL would encounter duplicate content. Backlinks pointing to the wrong version would also remain diluted. The 301 solves both problems at once: it unifies user flow and concentrates SEO equity.

Does this rule apply differently depending on the CMS?

WordPress, Shopify, or Prestashop handle this redirect differently. WordPress often enforces the choice right from installation through general settings, but many hosts configure an automatic redirect at the Apache or Nginx level without the user being aware. Shopify systematically forces either the www or non-www version depending on the initial DNS setup.

The problem arises during migrations or changes of hosts. A migrated site may lose its redirect rules if the .htaccess file or Nginx configuration is not transferred correctly. As a result, both versions become accessible again, and Google reindexes duplicate content. Checking this rule after every infrastructure change is crucial.

• Choose a canonical version (www or non-www) and stick to it permanently
• Implement a permanent 301 redirect on the server for all URLs of the unchosen version
• Check for consistency in Google Search Console by declaring the preferred domain
• Audit incoming backlinks to identify those pointing to the non-canonical version
• Test the redirect on all pages, not just the homepage, as some poorly written .htaccess rules may only redirect the root

Is this Google directive still relevant with widespread HTTPS?

Absolutely. The advent of HTTPS has even complicated the issue by multiplying the possible variants: http://example.com, http://www.example.com, https://example.com, https://www.example.com. Four distinct URLs for the same content. Many sites only correctly redirect three of these versions, leaving a gap through which Google can index duplicates.

I have seen sites lose 15-20% of their organic traffic after a HTTPS migration simply because www/non-www redirects were not replicated for both protocols. Google was indexing the non-www HTTP version (old backlinks) and the www HTTPS version (new version) simultaneously, diluting the signals. [To check]: the speed of signal transfer varies greatly depending on domain authority, from 2 weeks to 6 months.

Which version should you choose: www or non-www?

Google claims that the choice is indifferent, but practice nuances this statement. Non-www domains are slightly shorter (better for branding, fewer characters in URLs) and are the modern trend. Major tech brands (Apple, Tesla, Stripe) prefer this form.

www domains have a historical technical advantage: they allow for more flexible DNS management (separate cookies for subdomains, CDN configured on a CNAME www). For internationally scaled sites with complex infrastructure, this flexibility remains valuable. However, for 90% of sites, this argument no longer holds. Choose based on your branding consistency, not based on a supposed SEO advantage.

What happens if you change the canonical version after years?

Technically feasible, but risky. Changing from www to non-www (or vice versa) after 5 years of history equates to a partial domain migration. Google must recrawl all URLs, transfer signals, and update its index. During this transition period, ranking fluctuations are inevitable.

I assisted a client who switched from non-www to www after 8 years in existence. Despite perfect 301 redirects, the complete transfer of signals took 4 months, resulting in a temporary loss of 12% of organic traffic. Google maintained both versions in the index for 6 weeks before consolidating. Unless there is a major technical imperative, keep your current version if your site already has some age.

How can you correctly implement the 301 redirect?

On Apache (.htaccess), the rule must capture all HTTP/HTTPS variants and redirect to the canonical version. A common mistake is writing a rule that only redirects the homepage, leaving deeper pages accessible in both versions. Use regular expressions that capture the entirety of the path.

On Nginx, the logic is similar but the syntax differs. Many shared hosts do not grant access to the Nginx configuration file, forcing you to use their proprietary interface (cPanel, Plesk). Ensure the interface offers a "permanent redirect" option rather than "alias" or "DNS pointing," which do not generate a 301.

What mistakes should you avoid during configuration?

The redirect loop: if your rule redirects www to non-www, but another rule (WordPress plugin, CDN) does the opposite, the server returns a 500 error or too many redirects. Always test with curl -I to see the complete redirect chain. Only one 301 should appear, not two or three in cascade.

Another pitfall: temporary 302 redirects instead of permanent 301 ones. Some poorly configured hosts or CMSs use 302 by default. Google does not transfer signals on a 302, considering the change to be temporary. As a result, after months, both versions remain indexed, and your backlinks never consolidate.

How can you check if everything is working correctly?

In Google Search Console, declare your preferred domain (domain property or URL prefix based on your choice). Monitor the Coverage tab to detect URLs indexed in the non-canonical version. If non-www pages appear when you have chosen www, your redirect isn't applied everywhere.

Run a Screaming Frog crawl in "list" mode with both versions of 50-100 internal URLs. If SF returns 200 codes for both versions, your redirect is failing. Also, manually test URLs with parameters (UTM, session ID): some .htaccess rules forget to preserve query strings, breaking analytics tracking.

• Audit the 4 variants (http/https + www/non-www) with curl -I to check status codes
• Test deep URLs, not just the homepage, to ensure the rule applies globally
• Check the preservation of query strings in the redirect (utm_source, session parameters)
• Configure Google Search Console with the chosen canonical version and monitor for indexing errors
• Update major backlinks by contacting referring sites to point directly to the canonical version
• Document the choice in an internal wiki to avoid a future service provider breaking the configuration during a migration