Subdomain vs Subdirectory

TL;DR — A subdomain is blog.example.com (a prefix before your domain). A subdirectory (or subfolder) is example.com/blog/ (a path after it). You’ll hear that subdirectories are “better for SEO.” They’re not, inherently — Google says either can work. Subdirectories are often easier to run as one application and analytics scope, so unless you have a real reason to split things out, keeping everything together is the simpler choice.

What’s the difference?

It comes down to where the extra word goes relative to your domain:

• Subdomain — the word comes before your domain, joined by a dot: blog.example.com, shop.example.com, support.example.com. Technically this is a separate address, so it can live on a different server or run a different system (CMS).
• Subdirectory (also called a subfolder) — the word comes after your domain, joined by a slash: example.com/blog/, example.com/shop/. It’s always part of the same address as your main site.

Does it matter for SEO?

Here’s the honest answer most articles won’t give you: not as much as people claim. Google has said for years that it has no preference. Both can work fine.

The reason you see headlines like “we moved to a subfolder and traffic jumped 40%” is that those moves almost never happen on their own. When a team migrates a blog, they usually also improve internal links, delete weak pages, and upgrade the platform — all at the same time. Those changes are what move traffic. The URL structure mostly went along for the ride. (I dig into this in the Advanced tab.)

So which should you pick?

A simple rule of thumb:

• Use a subdirectory for things closely tied to your main site — a blog, a help center, docs. It keeps routing, analytics, and content management under one roof. A Search Console Domain property can aggregate subdomains too, so property count is not the deciding factor.
• Use a subdomain when something is genuinely separate — a different product, a different team or server, or country-specific versions of your site. There’s nothing wrong with a subdomain; it just takes a little more setup to manage.

The thing most people get wrong

Don’t migrate a working subdomain to a subdirectory just because someone told you subfolders rank better. If your subdomain blog is doing well and is well-linked from your main site, moving it is risk with little reward. Migrations always carry the chance of losing rankings during the transition. Want the full reasoning, the case studies, and what Google and Bing actually say? Switch to the Advanced tab.

TL;DR — Subdomains and subdirectories are both viable for search, but they are operationally different. Google evaluates per site whether a subdomain is treated as part of your site or as a separate entity — the deciding factor is integration (internal links, consistent UX, related content), not the dot-vs-slash. The case studies showing big gains from subdomain→subdirectory moves are almost always confounded: better internal linking, content pruning, a CMS upgrade, or analytics changes rode along. My position: don’t migrate for URL structure alone — the risk usually outweighs the upside. Subdirectories can carry operational advantages (shared routing, analytics, and publishing systems), but those are convenience wins, not a ranking edge. Bing’s concern is subdomain abuse (leasing, PBNs), not legitimate subdomain use. And per Google’s November 2024 site-reputation-abuse policy, moving content between structures doesn’t fix a spam violation.

What they actually are

A subdomain is a prefix before the root domain, separated by a dot — blog.example.com. It’s technically a separate hostname, so it can point at a different server, run a different CMS, and be owned by a different team. A subdirectory (subfolder) is a path segment after the domain, separated by a slash — example.com/blog/. It’s always the same hostname, and so the same “site” from a crawling and analytics standpoint. There’s a third option for international work — a ccTLD like example.de — which is always treated as a separate site.

What Google actually says (and what it means)

Google’s official documentation is blunt: “You should choose whatever is easiest for you to organize and manage. From an indexing and ranking perspective, Google doesn’t have a preference.” That’s the headline — no blanket advantage either way.

The nuance is more important than the headline. John Mueller has said, “In general, we see these the same,” but added that he’d “personally try to keep things together as much as possible.” The key word is personally — that’s a preference for operational simplicity, not a statement that subdirectories rank better.

The deepest point Mueller has made is that there’s no blanket rule at all: “sometimes that can include sub-domains, sometimes that doesn’t include sub-domains. Sometimes that includes sub-directories, and sometimes that doesn’t include specific sub-directories.” In other words, Google decides per site whether a subdomain (or even a subdirectory) is treated as part of the main site. The signals it uses are integration signals: internal linking between the section and the root, consistent branding and UX, thematically related content, and a coherent sitemap structure.

Gary Illyes confirmed the flip side at Pubcon 2023: subdirectories can be treated as separate sites too — “It’s not common on the English web but it is in other languages.” So the structural choice isn’t the lever people think it is; the integration is.

My position: the “wins” are usually confounded

I wrote the contrarian take on this for Ahrefs in March 2021, and I’ll restate it here because it still holds. The belief that subdirectories are SEO-superior rests on case studies that move a section from a subdomain to a subfolder and report a traffic jump. The problem is that those migrations almost never change only the URL.

When teams do these moves, they typically also:

• Improve internal linking — the section gets properly woven into the main nav and contextual links for the first time.
• Prune or rewrite thin content as part of the migration.
• Upgrade the CMS or platform, which often improves speed, templates, and crawlability.
• Fix analytics, so traffic that was always there is finally measured under one property (cross-domain/cross-host tracking gaps are common on subdomains).

Any one of those can move traffic. Attribute the whole gain to the dot-becoming-a-slash and you’ve confused correlation with causation. As I put it in the article: changes introduce risk, so think twice before changing from a subdomain to a subdirectory if the only reason you’re doing it is for SEO.

That doesn’t mean no case study isolates the structure. Pink Cake Box is the cleanest one I know of — Archive.org checks show no design or content changes accompanied the move, and organic sessions rose roughly 40%. The reverse test, IWantMyName moving subdirectory→subdomain and dropping about 47%, adds weight. I take those seriously. But they’re the exceptions that get clean measurement; the HotPads/HubSpot-style cases everyone cites had simultaneous changes, and the people who ran them (John Doherty on HotPads) have said as much. Correlation studies — the “11.8M results” and similar — are correlational by nature; sites that use subdirectories also tend to be larger and more SEO-mature.

Real technical differences (these are not myths)

Set the ranking debate aside — there are concrete operational differences, and they usually favor subdirectories for closely-related content:

• Search Console. A subdomain needs separate verification and its own property. (A GSC Domain property covers all subdomains at once, which softens this.) Mueller: “We do have to learn how to crawl them separately but for the most part that’s just a formality for the first few days.”
• Crawling. With a subdirectory, Googlebot already understands it’s the same server — no re-learning period. Mueller framed the subdirectory advantage as: “this helps us with crawling since we understand that everything is on the same server and can crawl it in a similar way.”
• Analytics. One GA4 property by default; no cross-domain/host configuration to capture the section’s traffic.
• Site names in search results. Google supports site names for domains and subdomains but not for subdirectories — a real, documented functional difference.
• Preferred Sources / property-level tools. Some Search Console surfaces operate at domain and subdomain level, not subdirectory level.
• Link equity. Backlinks to a subdirectory are unambiguously under the root domain. For a subdomain, whether that equity “flows” to the root depends on whether Google treats the subdomain as part of your site — the same per-site evaluation as everything else.

When a subdomain genuinely makes sense

Subdomains aren’t a mistake. Reach for one when:

• It’s a genuinely separate product or service — app.example.com, status.example.com, community.example.com.
• You need technical separation — a different CMS, a different stack, a different team, or different infrastructure that’s far easier to run on its own hostname.
• International targeting — Google lists subdomains (de.example.com), subdirectories (example.com/de/), and ccTLDs (example.de) as equally valid for geotargeting. Subdirectories are operationally simplest; ccTLDs send the strongest geo-signal; subdomains sit in the middle.
• Large enterprises with multiple business units, where separate GSC properties per team is a feature, not a bug.

What Bing cares about

Bing doesn’t prescribe one structure for legitimate sites. In “Some Thoughts on Website Boundaries” (Frederic Dubut and the Bing Webmaster Tools team, November 2019), Bing’s concern is inorganic site structure — PBNs, doorways, and subdomain/subfolder leasing, where third-party content is hung off a subdomain with minimal internal linking specifically to leak site-level signals to an unrelated party. Bing’s blunt warning: “If you offer free hosting on your subdomains and 95% of your subdomains are flagged as spam, Bing will expand penalties to the entire domain, even if the root website itself is not spam.” For a normal business using its own subdomain, none of this applies.

This connects to Google’s site reputation abuse policy update (November 2024), which is worth knowing: “Moving content to a subdirectory or subdomain within the same site’s domain name doesn’t resolve the underlying issue and may be viewed as an attempt to circumvent spam policy.” So you can’t structure your way out of a parasite-SEO problem.

The reverse-proxy workaround

If you’re technically locked into a subdomain — WordPress.com, a headless CMS, or a docs/help platform you can’t self-host — you don’t have to migrate to get a subdirectory URL. You can use a reverse proxy (Cloudflare Workers is the common route) to serve blog.example.com as example.com/blog/ from the crawler’s and user’s perspective. The URL everyone sees is the subdirectory path; the backend stays on the subdomain install. Cloudflare documents this pattern in a two-part Workers series. It’s a legitimate, widely-used technique — just be sure the proxied content is genuinely yours, not leased (see the abuse policies above).

Should you migrate? How to think about the risk

My decision rule:

• Stay put if your subdomain is established, ranking, well-linked from the root, and integrated. Migrating purely for URL structure is risk with little reward.
• Migrate when you’re doing a larger platform migration anyway, or the subdomain has real technical/performance problems — i.e., when the URL change is bundled with fixes that actually move the needle.
• If you do migrate, fix the confounders first so you can read the result cleanly: internal linking, content quality, redirects (301s), and analytics. Then the structure change is the only variable left.
• Follow Google’s official site-move workflow — map old URLs to new ones, redirect every one of them, update internal links and sitemaps, verify the new property in Search Console, and then monitor a representative set of URLs rather than assuming the move went cleanly.

Expect 2–4 weeks of fluctuation as Google processes redirects, with recovery typically inside 4–6 weeks when the migration is done correctly.

Where this fits

This is one piece of how you organize a site — siblings in this cluster cover URL structure conventions, site architecture and crawl depth, and faceted navigation, and the international angle overlaps with hreflang and ccTLD decisions in international SEO. The throughline is the same: structure is a way to organize and manage content, not a ranking lever in its own right.

AI summary

A condensed take on the Advanced version:

• Subdomain = blog.example.com (separate hostname). Subdirectory = example.com/blog/ (path on the same hostname). ccTLD = example.de (always a separate site).
• Google has no blanket preference. Official docs: choose what’s easiest to manage; “Google doesn’t have a preference” for indexing/ranking.
• Treatment is per-site. Mueller: sometimes subdomains are treated as part of the site, sometimes not — and the same for subdirectories. The deciding factor is integration (internal links, consistent UX, related content), not dot-vs-slash. Illyes confirmed subdirectories can be treated separately too.
• Patrick’s position: migration “wins” are usually confounded — better internal linking, content pruning, CMS upgrades, or fixed analytics ride along with the URL change. Don’t migrate for URL structure alone; the risk outweighs the upside.
• Cleanest case studies: Pink Cake Box (+~40%, no other changes) and the IWantMyName reverse move (−~47%). Most other cited cases had simultaneous changes.
• Real (non-ranking) differences: subdomains need separate GSC verification; site names in search work for subdomains but not subdirectories; subdirectories crawl without a re-learning period and use one analytics property.
• Bing targets subdomain abuse (leasing, PBNs), not legitimate subdomain use.
• Nov 2024 site reputation abuse: moving content between subdomain and subdirectory does not fix a spam violation.
• Reverse proxy (Cloudflare Workers) can serve a subdomain as a subdirectory URL if you’re technically locked into a subdomain.

Official documentation

Primary-source documentation from the search engines.

Google

• URL structure best practices — Google’s guidance on URLs; recommends ccTLDs or subdirectories for geotargeting, no stated preference for domestic content.
• Managing multi-regional and multilingual sites — the three equivalent URL options for international targeting (ccTLD, subdomain, subdirectory) with their pros and cons.
• Site names in Google Search — site names are supported for domains and subdomains but not subdirectories.
• Consolidate duplicate URLs — canonicalization methods (redirects, rel="canonical", sitemaps); relevant when migrating between structures.
• Site move with URL changes — Google’s official workflow for a URL-structure migration: mapping, redirects, Search Console verification, and monitoring.
• Site reputation abuse policy (November 2024) — moving content between a subdirectory and subdomain doesn’t resolve a spam violation.
• Subdomains and subdirectories (January 2008) — the oldest official Google blog treatment of the topic.

Bing / Microsoft

• Some Thoughts on Website Boundaries — Frederic Dubut and the Bing Webmaster Tools team on single-domain / subdomain / subdirectory models and the “inorganic site structure” penalty (PBNs, doorways, subdomain leasing).
• Bing Webmaster Guidelines — general guidance; no prescriptive preference between structures for standard sites.

Quotes from the source

On-the-record statements from Google and Bing. Where the source page supports it, each link is a deep link that jumps to the quoted passage.

Google — no preference

• “You should choose whatever is easiest for you to organize and manage. From an indexing and ranking perspective, Google doesn’t have a preference.” — Google Search Central documentation.

John Mueller, Google (Office Hours Hangout, May 25, 2018; reported by Search Engine Journal)

• “In general, we see these the same. I would personally try to keep things together as much as possible.” Coverage
• “So if it’s the same site then try to put them on the same site, essentially, and use subdomains where things are really kind of slightly different.” Coverage
• “There are lots of really strong opinions on this so, from my point of view, this is something that could go either way.” Coverage

John Mueller, Google (on per-site treatment)

• “sometimes that can include sub-domains, sometimes that doesn’t include sub-domains. Sometimes that includes sub-directories, and sometimes that doesn’t include specific sub-directories.”

John Mueller, Google (on crawling and Search Console, via Search Engine Journal)

• “this helps us with crawling since we understand that everything is on the same server and can crawl it in a similar way. Sometimes, it also makes it easier for users who recognize that these sections are all a part of the same bigger website.” Coverage
• “You’ll need to verify subdomains separately in Search Console… We do have to learn how to crawl them separately but for the most part that’s just a formality for the first few days.” Coverage

Gary Illyes, Google (Pubcon 2023, on whether subdirectories can be treated separately — cited in my Ahrefs article)

• “Yes. It’s not common on the English web but it is in other languages.” Source

Matt Cutts, Google (former head of webspam)

• “They are roughly the equivalent. I would basically go with whichever is easier for you in terms of configuration, your CMSs.” (2013)

Google — site reputation abuse (November 2024 policy update)

• “Moving content to a subdirectory or subdomain within the same site’s domain name doesn’t resolve the underlying issue and may be viewed as an attempt to circumvent spam policy.” Jump to quote

Bing — inorganic site structure (Frederic Dubut & Bing Webmaster Tools team, November 4, 2019)

• “If you offer free hosting on your subdomains and 95% of your subdomains are flagged as spam, Bing will expand penalties to the entire domain, even if the root website itself is not spam.” Jump to quote

Choosing (or migrating) a structure — checklist

Deciding where a new section should live

• Is the content closely related to the main site (blog, docs, help center)? → lean subdirectory.
• Does it need a separate CMS, server, team, or stack? → a subdomain is reasonable.
• Is it a genuinely distinct product or service (app., status.)? → subdomain.
• International targeting? → subdirectory, subdomain, and ccTLD are all valid; pick on operational cost vs. geo-signal strength.
• Do you need site names in search for this section? → only subdomains qualify, not subdirectories.

Before migrating subdomain → subdirectory (only if there’s a real reason)

• Confirm the move isn’t purely for SEO — if it is, reconsider.
• Fix internal linking, content quality, and analytics first, so you can read the result cleanly and aren’t just confounding the test.
• Map every old URL to its new path and set 301 redirects.
• Update internal links, sitemaps, and canonical tags to the new paths.
• Verify the new structure in Search Console (or use a Domain property).
• Don’t migrate to escape a site-reputation-abuse problem — it won’t fix it.
• Expect 2–4 weeks of fluctuation; monitor rankings and crawl stats through recovery (typically 4–6 weeks).

If you’re locked into a subdomain

• Consider a reverse proxy (e.g., Cloudflare Workers) to serve it as a subdirectory URL instead of a full migration.
• Make sure the proxied content is genuinely yours, not leased third-party content.

The mental models

1. Structure is organization, not a ranking lever. The dot-vs-slash choice is about how you organize and manage content. Google has no blanket preference. Start every conversation here — it kills 90% of the bad advice.

2. Treatment is per-site, decided by integration. Google decides per site whether a subdomain is “part of you” or separate. The signals are integration: internal linking, consistent UX, related content, coherent sitemaps. A well-integrated subdomain behaves like part of your site; a poorly-integrated subdirectory can behave like a separate one. Optimize the integration, not the URL.

3. The confounder test. Before you believe any “we moved and traffic jumped” story, ask: what else changed? Internal links, pruned content, new CMS, fixed analytics — if any moved at the same time, you can’t credit the URL structure. The only clean tests are the ones where nothing else changed (Pink Cake Box).

4. Convenience vs. capability. Subdirectories win on convenience (one GSC property, one analytics setup, no crawl re-learning). Subdomains win on capability (separate stacks/teams/servers, site names in search, clean separation for distinct products). Decide on which axis your situation actually lives.

5. The migration risk rule. Migrate only when the URL change is bundled with changes that genuinely move the needle, or when there’s a real technical problem. A working, well-integrated subdomain

• “subfolders rank better” is not a reason to migrate — it’s risk without reward.

Should this section use a subdomain or subdirectory?

Structure decisions that confuse cause and effect

Migrating only because subdirectories supposedly rank better

Google does not express a blanket ranking preference. Keep a healthy, integrated section where it is unless a real platform, product, or operating need justifies the risk of changing every URL.

Changing URLs, CMS, content, links, and analytics together

A bundled release makes the outcome impossible to attribute. Fix known content and linking problems first, document the baseline, and isolate changes where practical.

Treating a subdomain as unrelated in the site’s own navigation

Technical separation does not require user or link isolation. Give a legitimate subdomain clear contextual and navigational links so its relationship to the main site is understandable.

Check a migration sample from the command line

Put one old subdomain URL per line in old-urls.txt, then inspect the first response and destination on macOS or Linux:

while IFS= read -r url; do printf '%s\t' "$url"; curl -sS -o /dev/null -w '%{http_code}\t%{redirect_url}\n' "$url"; done < old-urls.txt

PowerShell equivalent:

Get-Content old-urls.txt | ForEach-Object { try { $r = Invoke-WebRequest -Uri $_ -MaximumRedirection 0 -SkipHttpErrorCheck; [pscustomobject]@{Url=$_; Status=$r.StatusCode; Location=$r.Headers.Location} } catch { [pscustomobject]@{Url=$_; Error=$_.Exception.Message} } }

The first hop should be a permanent redirect to the exact mapped subdirectory URL, not a home page or a chain. Run a full crawler and migration review before treating a sample as complete evidence.

Tools for a structure decision and migration

• SEO Migration Planner & Validator — build and review the old-to-new URL map, verify deployed redirects, and compare sitemap sets when a move is genuinely justified.
• Redirect Chain Mapper — trace whether an old subdomain URL reaches its matching subdirectory destination in one clean hop.
• Google Search Console — use a Domain property for a combined view and hostname-specific properties when teams need separate operational reporting.
• A link crawler — compare navigation, inlinks, canonicals, and crawl depth before and after the change so improved integration is not mistaken for a URL-format win.

Test yourself: Subdomain vs Subdirectory

Five quick questions on how subdomains and subdirectories really differ for SEO. Pick an answer for each, then check.

Resources worth your time

My related writing

• Subdomain vs Subdirectory: Subdirectories Are Not Better Than Subdomains For SEO — my full contrarian case, with the Mueller and Illyes quotes, the case-study confounders, and why I’d think twice before migrating for URL structure alone.
• The Beginner’s Guide to Technical SEO — where site structure decisions fit in the bigger picture.

My speaking

• How Search Works (SlideShare) — my walkthrough of crawling, rendering, indexing, and ranking; useful background for why “integration” matters more than the URL. (Standing disclaimer applies: “This is my understanding of systems… not going to be 100% complete or accurate.”)

From around the industry

• Google’s John Mueller On When To Use Subdomain vs. Directory (Search Engine Journal, Matt G. Southern) — the May 2018 Office Hours quotes in context.
• Subdomains vs. Subfolders: Which Is Better for SEO & Why? (Search Engine Journal, Roger Montti) — Mueller on crawling and Search Console treatment.
• Some Thoughts on Website Boundaries (Bing, Frederic Dubut) — Bing’s website-boundary models and the inorganic-site-structure penalty.
• Does Switching From a Subdomain to a Subfolder Improve SEO? Case Studies & Experts RoundUp (cognitiveSEO) — the best case-study roundup, honest about confounders.
• Subdirectories vs Subdomains: Which is Better for SEO? (Portent) — solid methodology and the Pink Cake Box / Archive.org verification.
• Subdomains vs Subdirectories Research: 20K Keywords Analysed (SE Ranking) — a transparent correlational study across 15 markets and 1.5M SERP positions.
• SEO Best Practices with Cloudflare Workers, Part 1: Subdomain vs. Subdirectory (Cloudflare) — the reverse-proxy workaround explained, with a Part 2 implementation guide.

Subdomain vs Subdirectory

A subdomain (blog.example.com) is a separate hostname; a subdirectory (example.com/blog/) is a path on the same hostname. Google has no blanket ranking preference — it decides per site whether a subdomain is treated as part of the site, based on integration signals.

Related: Website Structure, Canonicalization

Subdomain vs Subdirectory

A subdomain is a prefix added before the root domain, separated by a dot — blog.example.com. Technically it’s a separate hostname, so it can point to a different server, a different CMS, or a different team. A subdirectory (or subfolder) is a path segment after the domain, separated by a slash — example.com/blog/. It’s always part of the same hostname, and therefore the same “site” from a crawling and analytics standpoint.

The persistent myth is that subdirectories are SEO-superior. They aren’t, inherently. Google’s official line is that it has no preference, and that you should pick whatever is easiest to organize. The real nuance: Google evaluates per site whether a subdomain is treated as part of the main site or as a separate entity — and the same is true, less often, of subdirectories (Gary Illyes confirmed subdirectories can be treated separately too). The deciding signals are integration: internal linking, consistent branding, and thematically related content.

There are a few concrete functional differences (not ranking myths): subdomains require separate Google Search Console verification; site names in search results work for subdomains but not subdirectories; and Google’s November 2024 site-reputation-abuse policy makes clear that moving content between a subdomain and a subdirectory does not resolve a spam violation. ccTLDs (example.de) are a third option, used for international targeting and always treated as separate sites.

Related: Website Structure, Canonicalization

Changelog