YouTube Mentions Predict AI Visibility Better Than Backlinks, Ahrefs Finds
Ahrefs studied 75,000 brands and found YouTube mentions correlate with AI visibility more strongly than backlinks, Domain Rating or content volume. What the numbers say, and what they don't.
Most SEO teams still judge a brand's strength by links. Ahrefs tested that habit against AI answers, and the result is uncomfortable. In a study of 75,000 brands, the factor that tracked AI visibility most closely was not backlinks, Domain Rating or page count. It was how often the brand is mentioned on YouTube.
What Ahrefs measured
The study compared brand-level signals against visibility in ChatGPT, Google AI Mode and Google AI Overviews. "YouTube mentions" means any time a brand name appears in a video title, transcript or description. "YouTube mention impressions" weights those mentions by the views each video got.
The headline figures, as reported by The Next Web:
- YouTube mentions: a Spearman correlation of about 0.737 with AI visibility.
- YouTube mention impressions: about 0.717.
- Branded web mentions: roughly 0.66 to 0.71 depending on the platform.
- Domain Rating: about 0.27 to 0.33.
- Backlink count: around 0.19 to 0.22.
Ahrefs also reported a near-zero correlation between a page's word count and its chance of being cited in an AI Overview. That is a separate finding from the same research, and it fits the same story: volume of output matters less than being talked about.
Read the word "correlation" carefully
Spearman correlation tells you two rankings move together. It does not tell you YouTube causes the citations. Big brands get mentioned everywhere, so YouTube mentions may simply be a marker of brands that are already strong on other signals. Ahrefs says as much, and the coverage that repeats the number without that caveat is overselling it.
Still, the gap is wide enough to act on. A 0.74 against a 0.2 is not a rounding difference. The practical reading is that AI answers tend to name brands that exist in the wider conversation of the web, and video transcripts are a large, text-rich part of that conversation.
Why video transcripts are useful to a model
YouTube gives a retrieval system a lot of clean text: titles, descriptions, chapters and transcripts. A separate Search Engine Land analysis of 9 million AI answers across 400 plus enterprise brands found YouTube gave the strongest citation lift of any third-party source it tracked, at 2.8x, ahead of LinkedIn at 2.7x and Reddit at 2.4x. Two independent datasets pointing the same direction is worth more than one.
The same analysis found owned pages account for only 3% of citations on bottom-of-funnel commercial prompts, while 82% go to third parties. So the mentions that matter are mostly ones you don't control.
What to do with this
You can't buy a 0.737. You can do a few concrete things:
- Find out where your competitors get mentioned. If they show up in comparison videos and you don't, that is a gap you can name.
- Make your own videos answer real questions. Titles that match the prompt a buyer would type are more useful than branded teasers.
- Earn mentions in other people's videos. Creator partnerships and review coverage count for the mentions metric, not just your own channel.
- Keep doing link work, but stop treating it as the AI lever. Links still help crawling and classic rankings. They just don't explain AI visibility well.
Where tracking fits
The hard part is that YouTube's pull differs by engine and moves month to month. You need to see which videos are actually being cited for your prompts, and whether that changes after you publish.
This is where we'd put Promptwatch first. Its citation analytics break out YouTube citations alongside Reddit and page-level sources, and citation trends show how those sources move over time instead of giving you one snapshot. Prompt tracking then ties it back to the specific questions you care about. Most monitoring-only trackers stop at "were you mentioned", which leaves you guessing about why.
If you want to test the Ahrefs finding on your own category, start there: pull the YouTube citations for your top prompts, list the channels involved, and see whether you or your competitors are in them. Then decide whether video is a priority. Try Promptwatch for that first pass.