Quotes from the book «Technical SEO 2026»

2.6.1 YouTube: Impact of video content on SEO

2.6.1. 1. Video Rank and its relation to organic output

Video Rank is an aggregated internal assessment of the quality and relevance of video content that allows it to compete directly with traditional web pages in universal search rankings.

The merged engineering documents describing the Content Warehouse architecture showed that video doesn't exist in some siloed space; it's just another document type that requires a separate but integrated scoring system. We see mentions of the Video Content Search module, which confirms the existence of a specialized mechanism that processes and ranks video content.

It's about tracking behavioral signals on a scale that Google has publicly denied for web search, but has always applied to its own video giant: Play Rate (percentage launched), Average Watch Time, and, critically, Completion Rate. These are measures of user satisfaction with content, the same E-E-A-T, only through the lens of video. If a video is regularly launched and watched, that content gets a higher Video Rank, and NavBoost pushes it up, even in the main SERP, next to text documents.

Many SEOs waste time trying to optimize videos for keywords. Drop it. It's not technical optimization of the title, but rather psychological optimization of the first minute. If you don't hook the viewer in the first 60 seconds and deliver a high Play Rate and Average Watch Time, your entire Video Rank will plummet. Focus on the hook, not the tags.

2.6.1. 2. Links from YouTube description: RPR Transmission

Links placed in the description of a YouTube video can convey significant weight despite their formal nofollow status, as their evaluation mechanism is linked to the internal RawPageRank metric and the convention of passing link-weight.

RawPageRank (RPR) is, as the leak suggests, a raw, unadjusted PageRank value calculated by Google's algorithm based solely on raw link structure. It is the basis for more complex, cleaned-up versions such as PageRank-NearestSeeds (or pagerank_ns in the documentation). Herein lies the key: link-weighting, in the traditional sense, runs on a strict graph.

However, the leaks demonstrate that a link's ability to convey any ranking at all is conditional; user interaction, namely clicks, acts as a filter or gateway that determines whether a link is even considered. NavBoost tracks which links users actually click on, and the longer they stay on the landing page after clicking, the more valuable the signal. YouTube, as a first party and subsidiary of Google, is in a privileged position; data on clicks on links in video descriptions is available to Google with unprecedented accuracy.

Here's the bottom line: a link in the description of a popular, highly ranked (due to high Video Rank!) video with thousands of views and hundreds of clicks will pass this NavBoost click filter with flying colors. While a regular nofollow link from a low-visited blog can be completely ignored, a link from a YouTube channel that has high siteAuthority and is a trusted source in the eyes of the search engine doesn't just pass the «juice» but is, in fact, a direct, user-approved springboard for RPR in your direction. These aren't just links; they're user-validated links.

When Google says nofollow doesn't pass PageRank - that's technically true for RPR in its purest form, but in real ranking, where behavioral factors and NavBoost reign supreme, click-through traffic and the authority of the source being in Google's own ecosystem outweighs any attribute.

See also: my audio reports