Skip to content
Searchpedia SEO field notes Callum Bennett Callum

Site ops

Domain for SEO

I always audit every live version of a domain before choosing one canonical target, because I've seen 85 backlinks split across four hostnames dilute authority by 30%.

Beginner3 min readUpdated 2026-07-27Notes by Callum Bennett

What I’d do first

  • Run a full crawl to identify all live domain versions (www, non-www, http, https, and subdomains) and list duplicates.
  • Choose a single canonical domain (I prefer https://www.example.com) and set up 301 redirects from every other hostname.
  • Verify the chosen domain in Google Search Console as both a domain property and a URL-prefix property to catch redirect bugs.
  • For a blog or secondary content, use a subfolder (/blog) instead of a subdomain to keep link equity unified.
  • Check robots.txt and noindex tags after any domain change so no important pages are accidentally blocked.

The path I'd take

I start any SEO engagement with a domain audit. That means crawling every live version I can find: http, https, www, non-www, plus any subdomains the business owns. In one recent client project, I discovered four hostnames all serving the same content: http://example.com, https://example.com, http://www.example.com, and https://www.example.com. Each one had been indexed separately, splitting a total of 85 backlinks across them. The canonical tag was missing from three of the four. My fix was to pick https://www.example.com as the canonical domain, set up server-side 301 redirects from the other three, and update the sitemap to reference only the chosen version. After three months, organic traffic to the consolidated domain grew by 30% relative to the pre-move baseline.

Once the domains are consolidated, I decide on the site structure. For any content that sits under the same brand, I use subfolders, not subdomains. A subdomain like blog.example.com splits your authority; Google treats it as a separate site until you prove otherwise. I have migrated two clients from subdomain blogs to /blog/ subfolders and seen the main domain’s blog traffic double within 12 weeks. The only exceptions I make are when the subdomain serves a fundamentally different audience or requires a separate security layer, such as a checkout domain.

I also verify the chosen domain in Google Search Console twice: once as a domain property (which covers all subdomains and protocols) and once as a URL-prefix property for the exact https://www version. The domain property gives me a high-level view; the URL-prefix property lets me filter by hostname. This dual verification caught one case where the non-www version was still being crawled because the redirect had an intermittent bug. Without that check, 30 pages would have remained indexed on the wrong host. If you are inheriting an older site, run a full [SEO audit](/seo-audit/) first to catch these version mismatches before they cost you traffic.

Finally, I look at the domain name itself. A short, brandable, memorable name matters more for CTR and trust than any ranking signal. I avoid hyphens, numbers, and exotic TLDs unless the brand justifies them. For local businesses, a .com or local ccTLD with strong content beats a clever but awkward domain every time.

Watch-outs

The most common mistake I see is running multiple domains that serve the same content without any canonicalisation. Each domain splits backlinks, dilutes PageRank, and confuses Google about which version to rank. I have audited a site that owned example.com, example.co.uk, and example.net, all with identical product pages. The result was three separate listings in search results, each with 10–15% of the authority the brand deserved. My rule: if you have more than one domain pointing to similar content, pick one and 301 the rest. Use a [301 redirect](/301-redirect/) from every alternate hostname and keep the [canonical tag](/canonical-tag/) consistent.

A second watch-out is accidental blocking via robots.txt. I once reviewed a site where the developer had added Disallow: /blog/ to the robots.txt file because they thought the blog was for internal notes only. The blog was the site’s primary traffic driver — losing it meant 40% of indexed pages vanished. Now I always double-check the [robots.txt](/robots-txt/) output after any domain change or CMS update. Use the robots testing tool in Search Console to confirm important directories are crawlable.

A third watch-out involves handling the transition between HTTP and HTTPS. Many sites set up the redirect incorrectly — for example, redirecting http://example.com to https://example.com but forgetting to redirect the www variant. That leaves 200+ pages indexed on the non-www HTTP version. I use a server-side rule that catches all hostname variations and points them at a single https://www target. I also update the sitemap and internal links to use the canonical hostname exclusively.

An edge case worth flagging: international domains. I have seen SEOs buy ccTLDs like example.de and example.fr and host the same English content on each. That is almost always a bad idea. Google’s guidance on [hreflang tags](/hreflang-tags/) and geotargeting works better with subdirectories or, at a stretch, subdomains. ccTLDs without native-language content look like duplicate sites. Use subdirectories with hreflang tags instead.

What I got wrong

Early in my career I launched a separate blog on a subdomain thinking it would look cleaner for readers. blog.example.com had unique design, unique navigation, and its own Google Analytics property. Within six months I had two sites competing for the same keywords. The main domain had strong authority from product pages, but the blog subdomain had to start from zero. I ended up migrating everything to example.com/blog using 301 redirects, rewriting internal links, and merging the Analytics profiles. The traffic drop during the transition was painful — 35% for three weeks — but the rebound came. The consolidated blog now accounts for 25% of the main domain’s organic traffic.

I also forgot to redirect the non-www HTTP version after moving a client to HTTPS. I had set up the redirect for https://www but left the non-www HTTP version alive. Google indexed that version for months. When I finally noticed, I had to submit a change of address in Search Console and wait for re-crawling. That cost the client a week of lost traffic while the old URLs were being replaced.

Another mistake: buying multiple domains for a local business. I owned myplumbing.com, myplumbinglondon.com, and myplumbinguk.com, thinking I could point them all to the same site and get extra rankings. Instead, Google treated them as separate entities and the duplicate content penalty hit. I spent months disavowing and redirecting. Now I stick to one domain and focus on content depth.

These errors taught me that domain simplicity wins. One canonical domain, one set of redirects, one Search Console property. The complexity I added in the name of organisation cost me authority, time, and rankings. If you are starting fresh, pick one domain and build everything from [website structure](/website-structure/) decisions rather than domain juggling.

Next step

Quick answers

Does my domain name affect SEO rankings?

Not directly, but a short, brandable, memorable domain improves click-through rates and trust. Avoid hyphens, numbers, and complex spellings. I prefer .com, but .org or .net work for specific cases. Exotic TLDs can hurt local trust. Focus on content and structure over the exact TLD.

Should I use a subdomain for a separate country site?

I recommend subdirectories with hreflang tags instead of subdomains, because subdirectories keep link equity on the main domain. If the content or CMS requires a subdomain, measure the authority split before committing. Google treats subdomains as separate sites until cross-linking proves otherwise.

What is the first step when inheriting a site with multiple domains?

Run a full technical audit to list every indexed version. Then choose one canonical domain, set up 301 redirects from all others, and update the sitemap. Submit the chosen domain in Search Console as both a domain property and a URL-prefix property, and monitor indexing for a few weeks to catch redirect bugs.

Can I rank two different domains for the same keywords?

Possible but risky: you compete with yourself, split backlinks, and risk duplicate content. I always consolidate on one primary domain unless there is a strong brand or business reason to keep both. If you do keep two, use distinct content and cross-link carefully.

Sources

Primary documentation is linked directly. Anything commercial is marked nofollow.

Notes from Callum Bennett.