Are video subtitles really invisible to Google?

Why does Google ignore subtitles embedded in video players?

Search engines crawl the HTML code of pages, not the content of embedded media files. When you embed a video with subtitles via a player (YouTube, Vimeo, or even a custom HTML5 player), the subtitle files (.vtt, .srt) are dynamically loaded by the JavaScript player. Google does not parse these files to extract the text.

This technical limitation is not new, but it often goes unnoticed. Many professionals believe that simply adding a .vtt file in the tag of an HTML5 video is sufficient for indexing. This is false. The content remains invisible to crawlers, even though users can see it displayed perfectly.

What does Google mean by "declaring correctly" the page versions?

For a multilingual site offering videos with subtitles in multiple languages, each language must have its own URL. For instance: /course/video-intro-fr/ for French subtitles, /course/video-intro-en/ for English. On each page, the text of the subtitles must be present in the HTML DOM, either as a full transcription below the video or as a structured summary.

Then, you need to implement the hreflang tags to indicate to Google the language variants. Without this, Google may consider your pages as duplicate content or mis-target the language of results based on the user's geolocation.

Does this rule apply only to e-learning sites?

No. Google uses the e-learning example because these sites heavily rely on educational video content, but the rule applies to all types of sites that integrate videos. A corporate site with customer testimonials, media with reports, an e-commerce site with product demos: all are affected.

The nuance lies in the search intent. If your video is merely illustrative (a product demo), a complete transcription may not bring SEO value. However, if the video contains sought-after information (tutorial, training, expert interview), then the text of the subtitles becomes a major SEO asset.

• Video subtitles in a player are not indexed by Google, regardless of the format (.vtt, .srt) or player used
• For each language of subtitles, create a distinct URL with text visible in the HTML of the page
• Declare language variants with hreflang tags to avoid duplicate content issues
• The transcription must be in the DOM, accessible for crawling, ideally in structured text form (sections, timestamps if relevant)
• This logic applies to all sites using videos with informational intent, not just e-learning

Is this statement consistent with real-world observations?

Absolutely. I have audited dozens of sites with thousands of videos: those who simply load .vtt files via see no SEO traction on the keywords present in the subtitles. Search engines regularly test indexing rich content (PDFs, videos, audio), but to date, subtitles embedded in players remain a blind spot.

Google's confirmation is helpful because it settles a recurring debate. Some SEOs hoped that the evolution of JavaScript rendering would allow Googlebot to parse .vtt files. This is still not the case, and this statement indicates it is not on the short-term roadmap.

What nuances should be added to this recommendation?

Creating a page per language of subtitles can quickly become unmanageable if you have 50 videos and 10 languages: 500 URLs to maintain. For large catalogs, you need to make trade-offs. Prioritize main languages and videos with high organic traffic potential. Do not mechanically duplicate everything.

Another point: Google mentions "declaring correctly" without specifying if a simple language selector (via JavaScript) on a unique URL would suffice, provided the text content is indeed in the initial DOM. [To verify] with rendering tests, but the safest approach remains a distinct URL + hreflang.

Finally, pay attention to the quality of transcriptions. If you automatically generate text from the audio track (via Whisper, Google Speech-to-Text), proofread before publishing. A transcription full of mistakes or nonsensical content harms the user experience and could trigger negative quality signals.

In which cases does this rule not fully apply?

If your video is hosted on YouTube and you rely on YouTube SEO (internal YouTube search), then yes, the subtitles uploaded on YouTube do contribute to the internal ranking of the platform. But this does not change anything for classic Google web SEO: the web page embedding the YouTube video must still contain the text.

Another edge case: short videos (30 seconds, silent product demo). No dialogue, no subtitles, so the question doesn't arise. The same goes for purely decorative or emotional videos without informational content. No useful text = no indexing issue.

What concrete steps should you take if you have videos with subtitles?

First, audit your existing video pages. Use the Search Console "Inspect URL" tool and look at the rendered HTML: does the subtitle text appear? If not, you have a problem. Then, decide on the strategy: complete transcription below the video, or structured summary by chapters (if the video is long).

For a multilingual site, structure your URLs with a clear language parameter: /videos/video-title/fr/, /videos/video-title/en/. Implement hreflang tags in the of each variant. If you use a CMS (WordPress, Drupal), plugins like WPML or Polylang automatically manage this, but check the HTML output.

What mistakes should be avoided during implementation?

Do not create duplicate content by copy-pasting the same page with just the translated subtitles, without changing the URL or adding hreflang. Google will choose a canonical version at random, and you will lose traffic on other languages. Each language = unique URL + bidirectional hreflang.

Avoid also hiding the subtitle text in a closed accordion or a tab hidden by default. Google indexes this content, but it may deprioritize it if it is not immediately visible. Ideally, the transcription should be visible on scroll, under the video, in a well-identified block.

Another frequent mistake: forgetting the schema.org VideoObject markup. Add the transcript (URL of the transcription) or caption (URL of the .vtt file) properties in your JSON-LD. This does not index the text directly, but it helps Google understand the structure of the content and may unlock video rich snippets.

How to check if your site is compliant after changes?

Use Screaming Frog or Sitebulb to crawl your video pages. Check that each URL indeed contains indexable text (length > 300 words for complete transcription). Export the hreflang tags and ensure they form a complete cluster (each language points to all others + itself).

In Search Console, enable the "Video Coverage" report (if available for your account). It indicates if Google detects your videos and any potential errors. Finally, conduct a manual search test with textual snippets from the subtitles: if Google does not find your page with an exact phrase from the transcription, it means the text is not indexed.

• Audit your video pages with the "Inspect URL" tool to check for the presence of text in the DOM
• Create a distinct URL for each language of subtitles, with a clear slug (/fr/, /en/, etc.)
• Implement bidirectional hreflang tags on all language variants
• Integrate the transcription into the HTML, visible or accessible on scroll, not hidden in a player
• Add the schema.org VideoObject markup with the transcript or caption properties
• Check the coherence of hreflang clusters with a crawler (Screaming Frog, Sitebulb)