Signal & Flow's new URL checker tests any web address against Google's own guidelines and returns a scored verdict in seconds. We ran it on 10 real UK small business websites and found the same structural problems appearing across almost all of them.
None of these businesses knew their URLs were working against them. These issues are invisible on a standard visual review, but Google reads your URL before it reads anything else. So does every visitor who sees it in a search result or shares it in a message.
1. Underscores instead of hyphens
This is the most consistently misunderstood URL rule. Google treats hyphens as word separators. A URL like /boiler-installation is read as two clear words: "boiler" and "installation." Underscores are different: /boiler_installation is read by Google as a single compound term: "boilerinstallation." A term no one searches for.
The fix: Check your URLs for underscores. Replace them with hyphens, set up 301 redirects from the old URLs to the new ones, and update any internal links. Our URL checker flags this instantly.
2. URLs too many levels deep
Every forward slash in a URL is a level of depth. /services/plumbing/emergency/boiler-repair is four levels deep. Six of the ten sites we checked had key pages buried five or six levels down.
Google's crawl budget is finite. Pages buried deep are crawled less frequently and given less weight. To a visitor, a six-level URL looks like something buried in a back-office filing system. It signals a site that wasn't designed for people.
The fix: Key pages should sit no more than two levels deep. /services/boiler-installation is ideal. Anything beyond four levels for an important page is worth restructuring.
3. Dynamic parameters on real pages
Three of the ten sites had pages with URLs containing query strings, like ?id=123&cat=4. This makes a page look like the output of a database query rather than a real destination. To a visitor, it's unreadable and impossible to remember.
The fix: Most modern platforms allow you to set clean permalink structures. Switch to these in your settings. For older custom sites, this may require developer input, but it's a one-time change with lasting benefits.
4. Dates baked into evergreen URLs
Dates in URLs make sense for news. For a guide that is just as useful today as it was years ago, the date only signals age. A visitor scanning search results sees /2021/04/ and might choose a newer-looking alternative, even if the content is identical.
The fix: In WordPress, go to Settings, then Permalinks, and switch to Post Name. This one change can improve click-through rates on older content without touching a word of the article.
5. Content overlap
This was our most significant finding for e-commerce sites. We ran our compare mode on a product page and a category page. Google encounters the same item through two different routes and has to decide which one matters more. Without a canonical signal, Google is silently arbitrating between them on every search.
The fix: Add a canonical tag to your product pages pointing to themselves as the definitive version. This is a single line of code in the page's <head> that tells Google: this is the page that counts.
6. HTTPS: still not universal
One of the ten sites was still serving older blog pages over HTTP. These showed Chrome's "Not Secure" warning to any visitor who landed on them. There is no longer any acceptable reason for an HTTP page: it hurts rankings and immediately kills trust.
The fix: Paste your most important URLs into your browser and check the padlock icon. Your hosting provider can usually resolve mixed-content issues quickly.
What your URL is actually saying
A URL is the first thing Google reads about your page. A long, parameter-laden address signals that a site was built without the visitor in mind. The fixes are usually simple: a hyphen, a redirect, or a single canonical tag. But you have to know the problem exists before you can fix it.