A while ago, I started noticing the same pattern again and again.
A website owner would launch an English version of their WordPress site, configure hreflang tags, submit updated sitemaps, and wait for rankings to improve.
A few weeks later, the same question would appear:
"Why isn't my English page ranking in Google?"
The first suspect was always hreflang.
Honestly, I don't think I've ever been part of a multilingual SEO discussion where hreflang wasn't blamed within the first few minutes.
And to be fair, that seems logical.
The issue appeared after launching a multilingual website, so the problem must be somewhere in the multilingual setup.
Except that, in many cases, it wasn't.
After auditing several multilingual WordPress websites over the last few years, I started seeing the same pattern. Hreflang was often implemented correctly. Google understood the language relationships between pages.
The real issue was usually hiding somewhere else.
The First Surprise: Canonical Tags
One of the most memorable cases involved an English service page that looked perfectly fine at first glance.
The content was translated.
The URL structure was correct.
The language switcher worked.
The hreflang implementation passed every validation tool I tried.
Yet the page barely appeared in search results.
At first, I was convinced the issue had to be somewhere inside the hreflang setup. After all, everything else seemed normal.
Then I noticed something unexpected.
The English page contained a canonical tag pointing to the original language version.
In other words, the website was sending Google two different messages at the same time.
One signal said:
This is the English version of the page.
The other said:
This is not the page you should prioritize.
The funny thing is that once you notice the conflict, the problem suddenly looks obvious.
Before that moment, it can be surprisingly easy to miss.
Since then, I rarely review hreflang without checking canonical tags first.
Translation Is Not Localization
Another issue became much more common after AI translation tools improved.
And honestly, modern translation tools are impressive.
You can translate an entire website in a matter of hours. The grammar is usually correct. The spelling is fine. The sentences make sense.
A few years ago, that would have felt almost impossible.
But something still feels off on many translated websites.
What caught my attention was that some English pages technically looked perfect while still feeling strangely disconnected from their audience.
The content was translated.
The intent wasn't.
Many pages still contained examples written for local customers, references that only made sense in the original market, or calls to action designed for users with completely different expectations.
From the website owner's perspective, the page was English.
From the visitor's perspective, it often felt translated.
That difference is difficult to measure, but I suspect it matters more than many people realize.
The Metadata Nobody Checked
This was probably the most surprising discovery.
Not because the issue was complicated.
Because it was hiding in plain sight.
The page was translated.
The navigation was translated.
The URLs were translated.
Even the images had been updated.
The title tags weren't.
Neither were the meta descriptions.
The website owner assumed everything had been localized correctly because every visible element looked fine.
Why would anyone suspect metadata?
A closer inspection revealed that the multilingual plugin wasn't translating SEO metadata automatically.
As a result, English pages still carried titles and descriptions written for a completely different audience.
It was one of those moments where you spend an hour looking for an advanced technical issue and eventually discover that the problem was sitting right in front of you all along.
Internal Linking Tells a Story
Another thing became clear during these audits.
Most multilingual websites have one mature section and one younger section.
The original language version usually contains years of content, internal links, case studies, service pages, and supporting articles.
The English section often consists of a handful of translated pages.
Google can see that.
Users can feel that.
And rankings often reflect that reality.
I kept seeing websites where English pages technically existed but felt isolated.
They had fewer internal links.
Less supporting content.
Fewer references from other pages.
In some cases, the only way to reach them was through the language switcher.
That may be enough for users.
It is not always enough to build authority.
Maybe the Problem Is Time
This is probably the least exciting explanation.
But sometimes it is also the correct one.
Not every ranking issue comes from a technical mistake.
Sometimes a multilingual section is simply new.
The pages are indexed.
The setup is correct.
Google understands the language relationships.
Nothing appears broken.
The website just hasn't built enough authority yet.
A few months ago, I was watching the performance of a newly published English article.
It entered Google's index quickly.
Much faster than similar pages would have a few years ago.
At first, almost nothing happened.
Then impressions started appearing.
Slowly.
A few here.
A few there.
Not enough to celebrate, but enough to suggest Google was beginning to test the page.
That was a useful reminder.
Indexing and ranking are not the same thing.
It's easy to forget that when monitoring new content.
What I Took Away From These Audits
The biggest lesson was surprisingly simple.
When an English page struggles to rank, hreflang should not automatically become the main suspect.
Sometimes the problem is canonical.
Sometimes it is localization.
Sometimes it is metadata.
Sometimes it is internal linking.
And sometimes nothing is actually broken.
The page simply needs more time to establish itself.
That doesn't make hreflang unimportant.
It simply means multilingual SEO is usually much bigger than hreflang alone.
At least that has been my experience so far.
I'm curious whether other developers, SEO specialists, or website owners have seen similar situations.
Have you ever spent hours troubleshooting hreflang only to discover that the real problem was somewhere else?
Originally published on PBB Design:
https://pbb.design/en/blog/dev/google-ignores-your-hreflang/
