In late July I got eleven Search Console warnings across client sites in under two hours. Seven were nothing. The other four traced back to two causes that no business owner would ever think to check.
If your site was showing up on Google and then stopped, the instinct is to assume you did something wrong, or that Google changed the rules. Usually it is neither. Usually it is something boring on the server that nobody looked at.
Here is exactly what happened on a set of real sites, because the specifics are more useful than general advice.
Search Console sends an email for almost any change in how a page is categorized, including good ones. Of eleven warnings, seven were Google confirming that redirects and canonical tags were working correctly. One was literally titled as a failure and was Google telling me a fix had succeeded.
The rule of thumb: "Page with redirect" and "Alternate page with proper canonical tag" are good news. The one that means something is broken is any message starting with "Duplicate."
When somebody edits your site, a careful developer makes a backup first. Often that backup gets saved right next to the original, named something like index.html.bak.
Here is the part that surprises people, including developers. The server serves that backup file as a real web page. Full content, HTTP 200, completely crawlable. And because it was copied before the last round of SEO work, it does not have the tags that tell Google which version is the real one.
So Google finds two identical copies of your homepage, neither of which claims to be the original, and has to guess. Sometimes it guesses the backup. On one site there was also a backup copy of the sitemap being served, which meant Google was reading a second list of URLs pointing at old pages.
The fix takes ten minutes: move backups out of the public folder entirely. The damage takes weeks to undo, because Google has to recrawl and change its mind.
Your site can usually be reached at several addresses. With www and without. With https and without. With and without index.html on the end. To a person those are the same site. To Google those are up to five different pages with identical content.
A small server file called .htaccess is what tells the server to send everyone to one canonical address. Four client sites had no such file at all. Every one of them answered on every variation with no redirect.
On one of them Google decided the real version was the insecure, non-https, www address. The consequence was blunt: zero pages indexed. The site was perfectly good and completely invisible.
site:yourdomain.com into Google. Count the results. Fewer than the number of pages you have means Google is not indexing everything.If any of those come back wrong, that is very likely why your traffic dropped, and it has nothing to do with your content.
Because the site looks completely fine. It loads, it is fast, the phone number works, the photos are there. Every check a normal person would run comes back clean. The problem is in a file you have never heard of, or in a stray backup nobody remembers making.
That is most of what technical SEO actually is. Not tricks and not keyword stuffing. Just making sure the machine reading your site can tell what it is looking at.
If you want me to run those checks on your site, send me the address. It takes about ten minutes and I will tell you what I find whether or not you hire me for anything.
More posts on the blog. If you want to know whether Google and AI assistants can currently find your business, there is a free ten minute check. Every price is on the pricing page.
I will look at it and tell you what I find, free, whether or not you hire me.
Mándame tu sitio y te digo qué encontré. Hablamos español.
Get a free look