What technical SEO breaks have in common is that nobody notices them on the day
they happen. The site opens, the design is intact, the client is happy. But what
the page's <head> or the server's response tells Google has changed. The
result shows up weeks later, when the traffic graph turns down, and by then
finding the cause costs far more than it would have on the day.
These five are the breaks agencies that manage client sites run into most. Each can be seen with a single check.
1. An accidental noindex
An SEO plugin setting left on from staging, or WordPress's "Discourage search
engines from indexing this site" box, reaches production. Google drops the page
on its next crawl. The noindex can sit in the page source or in an
X-Robots-Tag response header, and the second is invisible when you look at the
page.
Check: noindex checker.
2. A robots.txt that closes the whole site
Disallow: / under User-agent: *. Right for staging, a disaster in
production. Google stops crawling; pages may linger in the index, but their
content is no longer refreshed.
Check: robots.txt tester.
3. A canonical that points somewhere else
On a site moved from staging, the canonical can still point to the old domain. A canonical is a hint, but a wrong hint can move the page's signals to the wrong address. Pagination or filter parameters can also make every variant look like a separate page with its own canonical.
Check: canonical tag checker.
4. A missing title, description or h1
A small change in a shared template can empty the <title> or meta description
on every page of the site. On JavaScript-rendered sites the h1 is drawn later,
so it may not be in the HTML the server sends at all, and that alone is not
necessarily a bug. The thing to compare is the same page yesterday and today.
Check: the free site check shows title, description and h1 together.
5. A removed tracking tag
Not an SEO error as such, but it stops the SEO work from being measured: the GA4 or Google Tag Manager snippet is removed in a theme update and the report comes back empty at month end.
Check: GA4 and GTM tag checker.
One check is not enough
These five checks give you today's snapshot. But breaks arrive on any day, with a deploy, a plugin update or an edit by the client's own team. Search Console does show them, but only after Google has recrawled the page and refreshed the report, which usually means days later.
The way to close that gap is to run the same checks on a schedule and compare each run with the last good state. Crawlens does that for your client sites: it reads noindex, canonical, title, forms and tracking tags on every scan of the pages you watch, and tells you with a dated record when one changes. What that looks like: sample change report.