SEO-friendly URLs and links: how to structure a business website
Review URL structure and crawlable links when planning a business website using Google's published recommendations.
Two Google Search Central pages cover this: URL structure best practices and link best practices. We read both in full on 15 November 2026. Everything attributed to Google below comes from them. Our own design suggestions are marked.
1. Requirements for a crawlable URL structure
Google says that if URLs do not meet these, it will likely crawl the site inefficiently (including extremely high crawl rates) or not at all.
- Follow IETF STD 66. Reserved characters used as data rather than URL delimiters need appropriate percent-encoding.
- Do not use fragments to change content. Example: example.com/#/potatoes. Google generally does not support fragments. If JavaScript changes content, use the History API.
- Use common parameter encoding: = between key and value, & between parameters, and a character like a comma for several values of one key. Google's "not recommended" examples use colons and brackets, or commas for everything.
2. Making URLs easy to understand
- Descriptive words over ID numbers.
- Your audience's language. Use words they search with, including non-Latin scripts where relevant.
- Percent-encode non-ASCII characters in link href attributes.
- Hyphens, not underscores. Google explains underscores are often used to keep concepts together (like format_date in code) while hyphens signal a word boundary. Joined words (greendress) are also listed as not recommended.
- Fewer parameters. Trim ones that do not change content.
- URLs are case sensitive. Google treats /APPLE and /apple as different URLs. If your server treats them the same, convert to one case.
- Multi-regional sites: a country domain (example.de) or a country subdirectory (example.com/de/) are both listed as recommended.
3. Common URL problems
Google says complex URLs, especially with several parameters, can create huge numbers of URLs for the same or similar content, so Googlebot may use more bandwidth than needed or fail to index everything. Its listed causes:
- Additive filters (beach, then beach and fitness centre...) that multiply views of one list.
- Irrelevant parameters: referral parameters, sorting parameters, session IDs. Google suggests cookies instead of session IDs, and consider robots.txt to block problem URLs.
- Infinite calendars: add nofollow to links to dynamically generated future pages.
- Parent-relative links in the wrong place, which can create endless bogus URLs if the server does not return the right status for nonexistent pages. Use root-relative links.
Blocking crawling is not a removal or privacy control: a blocked URL can still appear in Search, and Google cannot read a noindex rule on a page it cannot crawl. To fix crawling of such URLs, Google suggests robots.txt for dynamic URLs like search results, calendars, ordering and filtering, and refers to its faceted navigation guide, which we have not read.
4. Links: make them crawlable
Google says it can generally crawl a link only if it is an anchor element with an href attribute. It lists these as parseable: an anchor with an href (absolute, root-relative or relative), including with an onclick or a class, and links inserted by JavaScript if they use that markup. It lists as not recommended, though Google may still try: routerLink attributes, a span with href, anchors with only onclick, and javascript: URLs in href. The URL must resolve to a real web address.
5. Anchor text
- Put descriptive text between the anchor tags. For image links, Google uses the img alt text, so write it. If the anchor is empty it may fall back to the title attribute.
- If JavaScript inserts the text, check the rendered HTML with the URL Inspection tool.
- Good anchor text is descriptive, reasonably concise and relevant. "Click here" and "read more" are Google's examples of too generic; a whole sentence as the link is too long.
- Do not stuff keywords (Google points to its spam policies), and do not chain links one after another without context.
6. Internal and external links
- Google says every page you care about should be linked from at least one other page on your site, and there is no ideal number of links per page ("if you think it's too much, then it probably is").
- Linking out is fine and citing sources can help show trustworthiness. Use nofollow when you do not trust the source, not on every external link. Use sponsored (or nofollow) for paid links, and ugc (or nofollow) for user-inserted links.
Applying it to a services site with a blog (our suggestion, not Google's rule)
| Page | Example URL |
|---|---|
| Service | /services/mobile-app-development/ |
| Blog index | /blog/ |
| Blog category | /blog/category/startup-guides/ |
| Article | /blog/mvp-vs-prototype-vs-poc/ |
Keep slugs short, lower case, hyphenated and stable. Link each article from its category and from related articles with specific anchor text. If a live URL must change, see our site migration and redirect guides.
Limits
URL and link structure help crawling and understanding; they do not replace useful content, and these pages make no promise about rankings. The faceted navigation guide and Google's spam policies were not read. We give no numbers for crawl budget.
By Omnistu.