Google is not indexing my site in multiple languages, why?

[diller0054]

The site itself is hitno.com

It has been translated into 9 languages, the total number of pages is about 6 million.

The situation in Google is approximately the following, it sees about 2 million, but does not want to index them. Everything that is currently in the index, it usually removes after a couple of days, including keywords. Moreover, Google indexes in fits and starts, it initially added us about 900 pages, then removed up to 280 at the moment and continues to remove. The same situation was with keywords, from 4 thousand, now about 1.6 thousand. Although when he added keywords, visits suddenly went up.

I have already reviewed and rechecked all technical errors hundreds of times, everything is correct, although perhaps my eyes are already blurred and I do not see point-blank.

The site itself is already old, it has been open for the index for about a year, maybe a little more. Headers and all meta were rewritten individually, including for other languages, alternate tags are also registered.

The site map was initially made via xhtml:link, but since the site was not indexed, I did it by trial and error and simply reassembled it, divided it into languages and other important parts (Main pages, languages, categories, products, etc.)

Maybe someone is very knowledgeable and can help me figure this out, I would be very grateful.

 

[dzuran]

Did you check the hreflang annotations are correctly implemented and consistent across all language versions ?

 

[diller0054]

Yes, I have a separate block in the head, with all the languages in the alternate tags, they all point to themselves and there is cross-linking between them.

Also on all the pages where I switch languages, there is a link to all the language versions

 

[harmananders]

The site itself is hitno.com

It has been translated into 9 languages, the total number of pages is about 6 million.

The situation in Google is approximately the following, it sees about 2 million, but does not want to index them. Everything that is currently in the index, it usually removes after a couple of days, including keywords. Moreover, Google indexes in fits and starts, it initially added us about 900 pages, then removed up to 280 at the moment and continues to remove. The same situation was with keywords, from 4 thousand, now about 1.6 thousand. Although when he added keywords, visits suddenly went up.

I have already reviewed and rechecked all technical errors hundreds of times, everything is correct, although perhaps my eyes are already blurred and I do not see point-blank.

The site itself is already old, it has been open for the index for about a year, maybe a little more. Headers and all meta were rewritten individually, including for other languages, alternate tags are also registered.

The site map was initially made via xhtml:link, but since the site was not indexed, I did it by trial and error and simply reassembled it, divided it into languages and other important parts (Main pages, languages, categories, products, etc.)

Maybe someone is very knowledgeable and can help me figure this out, I would be very grateful.

That definitely sounds frustrating. If everything’s technically solid, it could be a crawl budget issue or Google seeing too much similar content across languages. Try prioritizing your most important pages, build a few quality backlinks to key URLs, and monitor which ones stick in the index. Sometimes a few strong signals can shift things.

 

[Jameswann]

Google may be treating a lot of your translated content as low value or duplicate content if it seems too similar or auto generated. Even, if technically correct, sometimes google holds back on indexing until it sees real engagement or external trust signals. Maybe try focusing on boosting user activity or getting a few natural backlinks to some high-priority pages. Also, testing with smaller, high-quality language sections might help google crawl more confidently before scaling.

 

[Crypto_Man]

Your site currently has about six million auto‑translated pages which makes Google crawl too much and flags many pages as low‑value or duplicates. Instead of rebuilding giant sitemaps, pick your best pages, merge or canonicalize any duplicates, fix your hreflang tags, and split your sitemaps into small, language‑specific files with clear priority and update frequency.
Also, wse google search console’s coverage and url inspection tools plus your server logs to spot crawl errors or accidental noindex tags, then slowly resubmit batches of top pages to rebuild google’s trust.

 

[brooke17br]

If Google isn't indexing your multilingual website, there could be several reasons.
1. Incorrect hreflang Implementation
2. Lack of Language-Specific URLs

 

[diller0054]

Your site currently has about six million auto‑translated pages which makes Google crawl too much and flags many pages as low‑value or duplicates. Instead of rebuilding giant sitemaps, pick your best pages, merge or canonicalize any duplicates, fix your hreflang tags, and split your sitemaps into small, language‑specific files with clear priority and update frequency.
Also, wse google search console’s coverage and url inspection tools plus your server logs to spot crawl errors or accidental noindex tags, then slowly resubmit batches of top pages to rebuild google’s trust.

I didn't quite understand about canonization.

The site map was originally a regular one, it wasn't made for languages, but just for files with 5 thousand entries, and I put each language in the tags <xhtml:link rel="alternate" hreflang="ru-ru" href="" />, I recently decided to change this out of desperation.

The product description was translated via Google Translate API, and the title was generated separately for each language (it was generated, not translated), the same situation is with the title, description. Moreover, we checked it via a neural network analyzer, it turned out that all the titles were written by a person.

If we go back even further, then before the moment when we had titles written by real people, our supplier provided them to us, the situation was about the same.

And I don’t, including I can’t understand the logic, I have another site, on the same template, where at the moment the only product is the name and a placeholder as an image, but Google indexed this site, without a sitemap, literally in 3 days and they weigh 4,100 pages, and they still weigh, only the keys are added little by little.

 

[Crypto_Man]

I didn't quite understand about canonization.

The site map was originally a regular one, it wasn't made for languages, but just for files with 5 thousand entries, and I put each language in the tags <xhtml:link rel="alternate" hreflang="ru-ru" href="" />, I recently decided to change this out of desperation.

The product description was translated via Google Translate API, and the title was generated separately for each language (it was generated, not translated), the same situation is with the title, description. Moreover, we checked it via a neural network analyzer, it turned out that all the titles were written by a person.

If we go back even further, then before the moment when we had titles written by real people, our supplier provided them to us, the situation was about the same.

And I don’t, including I can’t understand the logic, I have another site, on the same template, where at the moment the only product is the name and a placeholder as an image, but Google indexed this site, without a sitemap, literally in 3 days and they weigh 4,100 pages, and they still weigh, only the keys are added little by little.

Think of “canonicalization” as telling google, “Here’s the one main url for this page, ignore the others pls!”.

Right now your sitemap just chops up 5k url chunks, but if you make several smaller sitemaps (one for each language) and list only those main urls, google will crawl much more smoothly. Machine‑translated pages often feel thin or repetitive, so concentrate on your strongest pages: either merge similar ones or point them to the main version, polish the translations you can, and then use search console’s coverage to fix any errors. Your other site indexed fast because it’s tiny and simple; once you give google clear, tidy signals on this bigger site, it’ll index it much better too.

 

[thebeatsdude]

I had a similar issue before. What helped me was making sure each language version had its own URL (like /en/, /fr/, etc.) and using the correct hreflang tags. Also, I submitted all versions separately in Search Console. It took a bit of time, but eventually Google started indexing them properly. Hope that helps!

 

[diller0054]

Think of “canonicalization” as telling google, “Here’s the one main url for this page, ignore the others pls!”.

Right now your sitemap just chops up 5k url chunks, but if you make several smaller sitemaps (one for each language) and list only those main urls, google will crawl much more smoothly. Machine‑translated pages often feel thin or repetitive, so concentrate on your strongest pages: either merge similar ones or point them to the main version, polish the translations you can, and then use search console’s coverage to fix any errors. Your other site indexed fast because it’s tiny and simple; once you give google clear, tidy signals on this bigger site, it’ll index it much better too.

Thank you, you give me hope that my problem will be solved.

 

[diller0054]

I had a similar issue before. What helped me was making sure each language version had its own URL (like /en/, /fr/, etc.) and using the correct hreflang tags. Also, I submitted all versions separately in Search Console. It took a bit of time, but eventually Google started indexing them properly. Hope that helps!

Thank you very much for your experience