You ship a new campaign page. Does it live at site.com/campaigns/ or at go.site.com? The answer sounds like plumbing, but it changes how quickly the page draws on the authority your domain already earned. Get it wrong and an identical page ranks like it came from a stranger.
A subfolder sits inside your domain. A subdomain reads more like a separate property. That single distinction decides whether a fresh page borrows your domain's standing on day one or starts building trust from a near-blank slate.
What Google actually says
Google's position has been consistent and worth quoting accurately, because it gets misrepresented constantly.
In general, we see these the same. I would personally try to keep things together as much as possible.John Mueller, Google Search Advocate, May 2018 office hours
Mueller said this in a May 2018 office-hours session, and Google has held the line since: there is no fixed ranking bonus for a subfolder, and the crawler handles either structure. His practical advice was to reserve subdomains for sections that are "really kind of slightly different," and otherwise keep content together. So the question becomes: if Google treats them similarly, why do practitioners keep reporting that subfolders win?
Google says the two structures rank similarly. Practitioners report subfolders performing better. These don't contradict each other. Google evaluates the page it finds. The structure changes how fast that page accumulates the link equity and trust signals it needs to rank. Same rules, different starting line.
Why the gap shows up in practice
A brand-new page has almost no signals of its own, so inherited authority is most of what it has. Place it in a subfolder and it draws on the links, age, and history your domain already built. Place it on a subdomain and search engines often treat it as a distinct site that has to earn standing the slow way.
The clearest documented case comes from G2.com. When the team moved blog content to a new learn.g2.com subdomain, it took Google roughly three to four months to build trust in it, despite the parent domain carrying a Domain Rating of 88 and over 750k monthly visits. Jakub Rudnik, who managed the migration, noted the lag would not have happened on a subfolder. Backlinko's analysis lands in the same place: subdirectories work best for most businesses because they consolidate authority, simplify content management, and report in one Search Console property.
Published migrations point the same direction, and it is worth naming them precisely:
- Monster moved
jobs.monster.co.uktomonster.co.uk/jobs/and reported a 116% jump in SISTRIX visibility, with the German property gaining similarly. - Pink Cake Box moved its blog from a subdomain into a subfolder in 2015 and reported a roughly 40% lift in organic traffic to the primary domain.
- Wirecutter folded
thewirecutter.comintonytimes.com/wirecutterin 2020; its referring backlinks reportedly grew from about 30k to 1.6 million and search traffic roughly doubled within the year.
Treat all of these as directional, not as clean experiments. A migration bundles 301 redirects, fresh internal links, renewed crawling, and sometimes a redesign; Wirecutter in particular was a full cross-domain consolidation onto one of the strongest news domains on the web, preceded by two years of SEO work, so structure was one variable among many. The honest read is not "subfolders add X%." It is that across very different sites, the move consistently goes the same way.
Practitioners and Google have openly disagreed
This is not settled-by-decree. After Google repeated that it sees the structures "the same," SEOs pushed back publicly, citing repeated before-and-after data where consolidating into a subfolder lifted rankings. Both sides can be right at once: Google's ranking systems do not award a structural bonus, but link equity, crawl frequency, and trust accrue to a domain as a whole, and a subfolder lets a new page tap that pool immediately while a subdomain has to bootstrap its own. The disagreement is really about mechanism, not direction.
The technical how: a subfolder served by another system
A common objection is practical: "our campaign tool or blog platform runs somewhere else, so it has to be a subdomain." It doesn't. A reverse proxy lets site.com/campaigns/ be a true subfolder even when the app lives on entirely different infrastructure. The browser and Google only ever see one host.
The mechanic is simple to state. A request hits site.com/campaigns/, your edge intercepts it, fetches the response from the other origin (say campaigns.vendor.com), and returns it under the original path. Common implementations:
- Cloudflare Workers: a Worker on a route like
site.com/campaigns/*fetches the upstream and streams it back. - Nginx: a
location /campaigns/ { proxy_pass https://upstream; }block. - Vercel / Fastly: a rewrite rule (
rewritesinvercel.json, VCL on Fastly) that maps the path to an external origin without changing the URL.
The pitfalls are real and worth respecting:
- Path collisions:
/campaigns/must not already be a real route on your main site, or requests fight over it. - Absolute URLs and assets: the upstream often emits links, CSS, and image
srcvalues that point back at its own host. Those need rewriting to the subfolder path, or the page leaks the real origin and breaks styling. - Canonical tags: the proxied page must self-canonicalize to
site.com/campaigns/..., not to the upstream host, or you hand Google a conflicting signal and undo the whole point. - Caching and redirects: edge caches must key on the public path, and any upstream redirect (including trailing-slash normalization) has to be rewritten so users never bounce to the hidden origin.
Done right, none of this is visible to a visitor or a crawler. Done carelessly, you get a subfolder URL that quietly behaves like a separate site anyway.
When a subdomain is actually the right call
Subdomains aren't a mistake. They're a tool for a different job, and forcing everything into a subfolder is its own error. Reach for a subdomain when:
- The content runs on a genuinely separate platform you don't want to proxy (a help center on a third-party tool, a status page, a community forum).
- You need operational or security isolation (a separate app, separate auth, blast-radius containment).
- The thing is a distinct brand or a regionally separate operation with its own team and content strategy.
In those cases the separation is a feature, not a tax. The mistake is using a subdomain by default for content that is really an extension of your core site.
A decision tree you can apply
- Is this page an extension of your core offering (campaign lander, geo page, offer page, blog)? Subfolder. This is the default and covers most growth-team pages.
- Does it run on a different system you can't move? You can still use a subfolder via reverse proxy. Only fall back to a subdomain if proxying isn't feasible or worth the operational cost.
- Is it a separate brand, a third-party platform, or something that needs isolation? Subdomain, deliberately.
- Still unsure? Default to the subfolder. Reversing later means 301s, re-indexing, and a temporary dip while Google re-evaluates. Putting the page in the right place the first time costs far less than fixing it.
The default we recommend
For the pages most growth teams generate, deploy to a subfolder of your primary domain. They draw on your authority, rank sooner, and compound the domain's standing instead of fragmenting it. Save the subdomain for content that is truly its own product.
Iterant deploys campaign pages under your own domain, and the subfolder path needs no DNS change, so the default lands where the evidence points. A subdomain works too, via a CNAME. The AI drafts and proposes; you approve before anything publishes.