Search Engines & SEO

URL Structure

Definition

Every URL has fixed parts: scheme, host, path, an optional query string and an optional fragment. Structure is mostly about the path. A good path is short, lower case, built from real words and separated by hyphens — /guides/password-protect-a-pdf rather than /p?id=4471 or /Guides/Password_Protect_A_PDF. Google has said for years that hyphens are read as word separators and underscores are not, so password_protect is treated as a single unfamiliar token. Depth matters less than folklore suggests, but a page sitting five folders down is usually a sign the site grew without a plan. The path is shown in the results listing above or under the title, so it also does a small job of telling a searcher where they are about to land.

Why It Matters

Addresses are promises. Once a URL is published it lives in bookmarks, emails, chat messages and other people's links, and changing it breaks every one of them unless you leave a 301 redirect behind. A readable path also survives being pasted somewhere the title is not shown: a recipient can tell what /2024/annual-report is and cannot tell what /d/9f2c11. The keyword value of words in the path is real but small — the better reason to write clean paths is that people read them, share them, and decide whether to click based on what they see.

How It Works

Each path segment is either a directory and file on the server or a route the server maps to one. Keep segments lower case, because on a Linux web server /About and /about are two different addresses and serving both is a common source of duplicate content. Stick to ASCII where you can: non-ASCII characters are percent-encoded, and a path full of %C3%A9 is unreadable the moment it is pasted into a message. Choose one canonical form — with or without a trailing slash, with or without www — and redirect the other permanently rather than answering on both. When a page genuinely must move, keep the old address responding with a 301 for as long as links to it survive, which in practice means indefinitely.

Real-World Example

A consultancy publishes a report at q3-market-report.99helpers.site with the PDF at /q3-market-report.pdf. Six months later someone wants it renamed to a dated scheme. Because the original link went out in a newsletter to 4,000 people, they leave the original address serving the file and publish the new name alongside it instead of renaming anything. The subdomain and the file name are both chosen at upload time, which makes the url slug worth a minute's thought before pressing publish rather than a month's cleanup afterwards.

Common Mistakes

  • Joining words with underscores — they are not read as separators, so two clear words become one token nobody searches for
  • Renaming a published address for tidiness — every existing link breaks unless a permanent redirect is left in its place
  • Baking dates or version numbers into a path you intend to keep updating — /2024-pricing is wrong in January and correcting it costs you the links

Related Terms

Put a file online in seconds

Drop in a document, an image, a page or a whole static website and share the link — free, with no build step and no server to set up.

Host a file free →