SEO-Friendly URLs: 2026 URL Structure Guide
A URL is the part of a page that people see before they read a word of it. It sits in the address bar, travels with every shared link and appears in search results. It is also the part of a page that is hardest to change later, once other sites and your own menus point to it.
We wrote this guide for the people who make URL decisions: site owners, writers, developers and store teams across the United States, Europe and South Korea, whether they are planning a new site, cleaning up an old one or settling a slug debate. It covers what your URL structure needs for crawling, what Google only recommends, which popular rules are myths, and when to leave a working URL alone. Each rule names its Google source, and each opinion is marked as ours.
What makes a URL structure SEO-friendly in 2026?
An SEO-friendly URL structure does two jobs. First, it meets the few hard rules that let Googlebot crawl your pages without wasted effort. Second, it follows Google's softer advice to use plain, readable words that people can grasp at a glance. The hard rules protect crawling, while the soft advice serves the people who see and share your links.
Google's URL structure guide lists its requirements first and warns that missing them can make Google crawl a site badly, with very high crawl rates or no crawling at all. The guide says reserved characters must be percent encoded, the part after the hash sign should not change what a page shows, and parameters should follow a common format.
After that, the guide turns to advice and suggests building URLs in a way that is "most intelligible to humans", and it says, when possible, to "use readable words rather than long ID numbers in your URLs". Google's SEO Starter Guide offers example.com/pets/cats.html as a URL that helps users, and our on-page SEO checklist covers the other page basics.
Which URL choices help, and which ones hurt?
Readable words and hyphens between them help both people and crawlers. Mixed case, session IDs and content loaded behind a fragment do harm, because each one can split a page into several URLs or hide content from Google, and the table below sets Google's advice next to ours.
The middle column follows Google's URL structure guide, and the last column is our own view.
| URL choice | What Google says | Our view |
|---|---|---|
| Readable words | Recommended over long ID numbers | Write for the person reading the link |
| Hyphens between words | Recommended over underscores | Make them the default for every new slug |
| Mixed upper and lower case | Case variants count as different URLs | Pick lowercase and redirect the other forms |
| Session IDs in the URL | Avoid them and use cookies instead | Strip them out before launch |
| Content behind a fragment | Generally not supported, use the History API | Give each view its own real URL |
In our view, each of these choices gets harder to fix once pages are live and linked, and that is the best reason to settle them before launch.
Do keywords in a URL still help rankings?
Hardly at all, according to Google, whose SEO Starter Guide says keywords in the domain or URL path alone do little for ranking beyond showing up in breadcrumbs. Words in a URL still matter for people, because a clear path tells a reader what the page holds before they click on it.
The exact wording in Google's SEO Starter Guide is that "the keywords in the name of the domain (or URL path) alone have hardly any effect beyond appearing in breadcrumbs". The same guide says the top-level domain matters only when you target users in one country, and even then it calls that choice "usually a low impact signal".
Our take: a URL matters most for crawling and for people, and much less for ranking. Get the URL structure right on day one, keep it stable, and do not rewrite working URLs only to squeeze in a keyword. Use keywords in a slug because they help a reader trust the link before clicking, and pick those words the way you pick them for the page, starting from our keyword research guide.
Should words be split with hyphens, underscores or nothing?
Use hyphens, as Google's URL structure guide recommends them over underscores between words and also advises against joining words with no separator at all. A hyphen is the one choice that reads cleanly for people and lets search engines see where one word ends and the next begins.
Google gives its reason in one line, saying it recommends hyphens because "it helps users and search engines better identify concepts in the URL", while underscores, it says, are commonly used to hold concepts together, as in function names in code. In example.com/summer-clothing you see two words, while example.com/summer_clothing looks like one joined term, and the guide also lists example.com/greendress as a pattern it does not recommend.
We think this is the easiest rule for SEO-friendly URLs to get right, and it works the same way in slugs and folder names.
Why do capital letters and stray variants create duplicate URLs?
Google treats URLs as case sensitive, which means a page reached through two spellings can count as two URLs with their own content, and we treat small variants, such as a trailing slash on some links and not others, as the same kind of risk. Each extra version is a duplicate that can waste crawling.
Google's URL structure guide says Google treats example.com/APPLE and example.com/apple as two different URLs, and if your server shows the same page for both, the guide says to convert all text to one case.
Google's SEO Starter Guide adds that duplicate content is not a spam violation, but it can waste crawling, and its advice is to redirect the extra URLs to the preferred one. When you cannot redirect, use a canonical link instead, as our canonical tags guide explains.
In our view, pick one form for letter case and one for the trailing slash, then send the other form to it with a permanent redirect, and our HTTP status codes guide explains which code each redirect should return.
How should folders mirror your site structure?
Folders should group pages on the same topic, the way aisles group goods in a store. Google's SEO Starter Guide says that on sites with more than a few thousand URLs, grouping similar topics in folders can help Google learn how often each folder changes, while on smaller sites, in our view, the main gain is a clearer path for readers.
The guide's example is a policies folder that rarely changes next to a promotions folder that changes often, and Google can learn that pattern and crawl each folder at its own pace. For example, an outdoor gear shop in Portland, Oregon might run product URLs made of long ID strings, while a path like example.com/camping/tents would tell a shopper where the page sits before the click.
In our view, two or three levels is plenty for a typical site, a human rule rather than a Google one. Our site architecture guide shows why folder depth and click depth differ, and our internal linking strategy covers the links behind click depth.
When do URL parameters turn into a crawl problem?
Parameters become a problem when they create many URLs for the same content. Google's URL structure guide warns that overly complex URLs can make Googlebot use far more bandwidth than needed and can keep Google from indexing all your content, and the causes it names include stacked filters, sorting options, referral tags and session IDs.
When you do need parameters, Google asks for a common format, with an equal sign between each key and its value and the ampersand character between one parameter and the next. Trim any parameter that does not change the content, and keep session IDs out of URLs by using cookies instead.
For problem URLs such as internal search results, the guide suggests a robots.txt block, which our robots.txt and noindex guide covers. Filter-heavy stores should read our faceted navigation guide, and our log file analysis guide shows where crawl time goes.
Why should the part after the hash sign never change the page?
Google Search generally does not support URL fragments, the part after the hash sign, and Google's URL guide says not to use fragments to change the content of a page. If JavaScript changes what the page shows, the guide points you to the History API instead, which can give each view a real URL.
This matters most on single-page apps, where a click swaps the content without loading a new page, and if each view lives behind a fragment, every view can end up sharing one URL in Google's eyes. In our view, any view you want people to find in search needs its own clean path, and our JavaScript SEO guide covers rendering for these sites. Filters you keep out of search are a separate case, as our faceted navigation guide shows.
How should URLs work for other languages and countries?
Give each language its own URL and use the words your audience uses, as Google's guide to multi-regional sites says to use separate URLs for each language version instead of cookies or browser settings. For country targeting, a country domain and a country folder both work, while that guide marks locale parameters "Not recommended".
Google's URL guide shows a country domain like example.de and a country folder like example.com/de/, while Google's multi-regional guide says a country domain gives clear targeting but costs more and covers one country. A subdomain is easy to set up, though users may not know if "de" means a language or a country, and the same doubt applies to a folder, which is easy and low upkeep but runs from one server location. Once the versions exist, our hreflang guide shows how to point Google to the right one.
Google's URL guide also says to use words in your audience's language, such as German words for German readers. For example, a garden shop in Munich, Germany can write German words with umlauts in its slugs. Its links should carry those letters in percent encoded form, which the guide asks for with any character outside the ASCII range. For example, a skincare brand in Seoul, South Korea can use Korean words in the slugs of its Korean pages, and we think that brand should also register its site with Naver, which runs its own crawler and index, as our Naver SEO guide shows.
Should you change old URLs to make them cleaner?
In our view, usually not, because a working URL has links, bookmarks and history attached to it, and moving it means a redirect, plus time for Google to recrawl and pass signals to the new address. Change an old URL only when it is broken, misleading or hard to read, not just to add a keyword.
Google's redirect guide says to use a permanent server-side redirect whenever possible to change how a URL shows in search, and the same guide says the 301 and 308 codes both mean a page has moved for good. Google's site move guide says to keep redirects as long as possible, generally at least one year, which gives Google time to transfer signals, recrawl, and reassign links from other sites. Moving paths within one domain needs no change of address request in Search Console.
We think a weak page needs a content refresh more than a new URL, and our content refresh guide makes the same case. Our redirects and site migration guide covers the mechanics of a move. Before you change a URL, run through this short check:
- Confirm the old URL is broken, misleading or hard to read
- Choose the new slug once and plan to keep it
- Set a permanent redirect from the old URL to the new one
- Update your own links to point straight to the new URL
- Keep the redirect live for at least one year
What does Google show of your URL in search results now?
On desktop, Google shows a visible URL with two parts, your domain and a breadcrumb trail, but on mobile that changed on January 23, 2025, when Google stopped showing breadcrumbs and cut the visible URL down to the domain alone. Your path still appears on desktop, which in our view keeps a clean URL structure worth the effort.
Google's visual elements gallery puts it plainly: "A visible URL has two parts: domain and breadcrumb". The breadcrumb shows where the page sits in the site's hierarchy. Google's Search Central blog post says breadcrumbs got cut off on small screens, which led Google to drop them from mobile results everywhere.
In our view, the change lowers the value of tidy, deep paths on mobile, not the value of a clean structure, and a clear site name now carries more of the load there. Our guide to site names and favicons shows how to state the name you want. For the wider mobile picture, see our mobile-first indexing guide.
A URL still makes its first impression in the address bar, long before anyone reads the page behind it. Readable words, hyphens and a stable path are what make SEO-friendly URLs clear at a glance. Once those choices are settled, your effort belongs on the page itself.
Frequently asked questions about SEO-friendly URLs
How long should a URL be?
Google's URL guide does not give a length limit. Its advice points toward readable words and fewer parameters, and both habits keep URLs fairly short. In our view, short and readable is a human rule rather than a Google rule, and a slug should name the topic in a few plain words and then stop. If you can read the URL aloud in one breath, it is short enough for a typical page on your site.
Should I remove small words like "and" or "the" from slugs?
Google's URL structure guide asks for readable words and does not address small words, which makes this our opinion. Dropping words like "the" or "a" keeps a path tight without losing meaning, but keep a small word when removing it changes the meaning or makes the path read like code. A path about summer clothing for kids, for instance, reads better with "for" left in. Aim for a path a person can read at a glance.
Is it fine to end URLs in .html?
Yes, and Google's SEO Starter Guide uses a page ending in .html as its own example of a helpful URL, with readable folder and page names. In our view, the ending does not make a URL harder to read, and the bigger issue is consistency. If some pages end in .html and others do not, links get mixed up and duplicates can creep in. Pick one pattern for new pages, and leave old pages alone, because each change means a redirect.
Are dates in blog URLs a problem?
Google's URL structure guide does not tell you to avoid dates in a path. In our view, dates cause a practical problem instead, because a year in the path can make an evergreen post look old once the year turns, and fixing that later means a new URL plus a redirect. For news and event pages, a date helps readers, but for guides you plan to update, decide on day one whether the date belongs in the path.
Should a blog sit on a subdomain or in a folder?
Google's SEO Starter Guide says to do whatever makes sense for your business, noting that folders may be easier to manage, while subdomains can suit some topics or industries, and it does not pick a winner. In our view, small teams will usually find a folder easier to run, because the blog shares one domain, one setup and one set of reports. A subdomain makes more sense when a separate team or system runs that part of the site.
How long should redirects stay after a URL change?
Google's site move guide says to keep redirects for as long as possible, generally at least one year. That window lets Google transfer signals to the new URLs, recrawl them, and reassign links from other sites that still point to the old ones. From the user's side, the same guide suggests keeping redirects with no end date and updating your own links to point straight to the new URLs, and in our view a working redirect rarely deserves removal.
Do tracking parameters hurt?
They can add clutter, because Google's URL guide lists referral parameters among the causes of needlessly high numbers of URLs for the same content, and each tagged link can look like a separate URL to a crawler. The SEO Starter Guide says to redirect extra URLs that show the same page, and to use a canonical link when you cannot redirect. Our advice is to keep tracking tags off your internal links and save them for campaign links from outside.
Is breadcrumb markup still worth adding?
Yes, for desktop results, because Google's January 23, 2025 post says it still supports breadcrumb markup there. The breadcrumb report in Search Console also continues, and you can still test markup in the Rich Results Test. The SEO Starter Guide adds that Google learns breadcrumbs from the words in your URL, and that markup can influence them. We think stores gain the most, and our product schema guide covers it alongside other product markup.