How subdomains and subfolders differ
A subfolder is a path on the same host: everything after the slash still belongs to one website, served by one server, under one name. A subdomain is a separate host name that the domain name system points wherever you like — it can run on different hardware, a different platform and a different technology stack from the main site. That flexibility is usually why teams reach for one.
Search engines can crawl and index either. Google has said consistently that it works out the relationship between a subdomain and its parent domain in most cases. The difference shows up in how reliably signals gather, not in whether the pages are eligible to rank at all. A subfolder sits inside the same house, so internal links, crawl paths and the reputation the domain has earned are pooled by default. With a subdomain, that pooling is something the search engine infers rather than something the structure guarantees.
Why the choice matters
Most of the cost is operational rather than mysterious. Each subdomain usually needs its own Search Console property, so reporting fragments. Cookies do not cross a host boundary unless configured to, so a visitor who moves from the main site to a booking subdomain can be counted as a new session from a new source — this is exactly the problem cross-domain tracking exists to solve. Consent state, tag containers and A/B testing tools all need the same treatment.
There is an editorial cost too. Content on a subdomain tends to be written by a different team, linked to less often from the main navigation, and forgotten in internal linking plans. The structure quietly encourages the neglect.
Where the decision goes wrong
The most expensive mistake is moving an established section without mapping it properly. Every old address needs a permanent redirect to its exact new equivalent, not to a category page and certainly not to the homepage, and internal links should be rewritten to point at the destination rather than bouncing through the redirect. Treat it as a site migration, because that is what it is.
The second is splitting a small site for internal reasons — one team owns the blog, another owns the shop — and ending up with several thin properties instead of one credible one. The third is country subdomains created without the language and region markup to tie them together, so the versions compete with each other. And the fourth is choosing a subdomain because a platform seemed to demand it, when a reverse proxy could have served the same software under a folder.
How to act on it
Default to a subfolder and make the subdomain justify itself. A subdomain earns its place when the section genuinely runs on software you cannot merge into the main site, or when it serves an audience that should stay separate, such as a careers portal or a customer login area.
If you keep a subdomain, close the gaps deliberately: add it as its own property, configure tracking across the boundary, and link to it from the main navigation and from relevant pages as you would any important section. Where the structure is already tangled, an audit under technical SEO is the sensible first step before anything is moved.
