The URL is the first thing I read when I open a new site. A bot reads it before the page even loads.
Most small business sites never choose that string on purpose. The platform generates it, nobody touches it, and it ends up a number or a folder path nobody picked.
That string is doing work whether anyone planned it or not.
I see the same handful of problems on almost every audit: a homepage with a clean address sitting next to blog posts, product pages, and service pages stuck with a platform's default numbering. A slug nobody would recognize. No plan for what happens when a page gets renamed. None of it is dramatic on its own. Together it adds up.
What a URL Actually Carries
A clean URL names the page in plain words. A page at /blog/drone-pricing-arizona tells a bot and a person the same thing before either one reads a sentence. A page at /blog/post?id=4821 tells them nothing.
Google has said for years that a descriptive address helps people and search engines understand a page faster. The same logic carries into AI search. A bot with limited time on a site reads the cheapest signals first. The URL is one of the cheapest.
It will not fix a thin page. It will not replace real content. It is one more piece of evidence, read before anything else.
The Folder Depth Problem
Folder depth is a separate issue. A page buried at /services/category/subcategory/page tells a bot that the page sits far from anything important. Depth reads as priority, whether that was the intent or not.
I flatten it when the page matters. A service page a buyer actually searches for should sit close to the root. Not nested under three layers of internal filing that made sense to whoever built the site and nobody since.
Most small business sites built on a template inherit this problem automatically, because the platform organizes pages by its own internal logic, categories, tags, and parent pages, rather than by what a buyer is trying to find. A site can have excellent pages that never show up anywhere near the top of a result because the address around them reads like an afterthought.
Parameters and Duplicate Pages
Tracking parameters and session IDs create a different problem. One page can end up living at ten addresses: ?utm_source=, ?ref=, ?sessionid=, each one technically a new URL pointing at the same content.
A canonical tag tells a bot which version is the real one. Our post on canonical tags covers that signal directly. Without it, a site can split its own authority across duplicates it never meant to create.
I check for this on every technical audit, before any content work starts.
The fix takes an afternoon on most sites. Not a rebuild.
What People Read Too
A person scanning a search result reads the URL before clicking. /menu reads differently than /page-id-00231. One looks like a business that knows what it is doing. The other looks like an accident waiting to happen.
The same string does double duty. It is a trust signal for a buyer and a legibility signal for a bot, read in the same half second, by two different readers who never compare notes.
The Fix Is Not a Rebuild
Fixing this does not mean rebuilding the site. I audit the full URL list, rename the pages worth keeping, and redirect the old address to the new one so nothing breaks on the way.
- Rename pages by what a buyer would actually search, not by the internal category name.
- Flatten folder depth on any page that matters for search or AI visibility.
- Strip tracking parameters from any address meant to be indexed or cited.
- Redirect every old address once, correctly, and check that it resolves.
Skip the redirect step and the fix becomes a new problem. A renamed page with no redirect is just a page that used to rank, now gone.
When a Slug Should Change
Not every URL is worth touching. A page already ranking, already cited, already working stays where it is. I don't touch it. Changing a working address is a real cost for a maybe gain.
The pages worth fixing are the ones already underperforming. The auto-generated slugs. The parameter strings. The pages nobody has looked at since launch. Start there, and stop once that list is clean.
I build a list before I touch anything: every page on the site, what it currently ranks for, what its address says, what its address should say, and whether the fix is worth the redirect it would require. The list takes longer to build than the fixes take to make.
A URL will never carry a business on its own. It is one line in a longer audit, worth fixing once and leaving alone after that. Cleaning it up is part of what the SEO / AEO / GEO Foundation covers before any content work starts.