YouTube is becoming an increasingly important source of information for Google’s large language models (LLMs), according to models’ citations tracked on the Evertune AI marketing platform.
YouTube’s share of citations has roughly tripled on Google AI Mode and Google AI Overview since October. It is trending upwards on Gemini, as well, while its share on Perplexity and ChatGPT has been more stable.
Of the five domains we examined - LinkedIn, Reddit, Wikipedia, Facebook and YouTube - YouTube has become the most cited on Google AI Mode and Google AI Overview.
YouTube’s ascendance in Google LLMs is a reminder that models are ever changing as model makers release new versions, some of which behave quite differently than their predecessors. Domains, too, can affect models’ sourcing behavior by altering their scraping permissions.
Reddit Still Relevant
In our February review of domains’ share of sources, the headline was clearly Perplexity’s enormous trust in Reddit. The LLM cited Reddit in 20% or more of its responses at times in January and February, based on a rolling average.
This is no longer the case. Reddit’s rolling five-day average of weekday prompts on Perplexity is now down to 5%. A five-day average of weekdays was used instead of a rolling seven-day average given lower volumes of prompts made on weekends.
Still, Reddit had the highest share of citations on Perplexity of the five domains we reviewed. Its share on Google AI Overview has increased every month since November, and it has seen a meaningful rise on Gemini as well.
Model Citations in Context
While observing trends in where models source their information can help inform a Generative Engine Optimization (GEO) strategy, these trends come with some important caveats.
First, no one publisher or domain dominates any model’s sources. A handful of domains tend to lead any topic’s source distribution, but there is always a long tail of thousands of other sources.
Second, not all prompts trigger an internet search. Sometimes the model is able to provide an answer using only the foundational knowledge it learned during training.
Lastly, just because a model cites a URL doesn’t mean that URL actually informed the response. We expose this in the Evertune platform with the Topic Relevance metric, which measures how much a source influences AI's understanding of a category and topic.
Methodology
At Evertune Research, we track hundreds of brands across 250 categories. For this analysis, we reviewed weekday prompts from September 25, 2025 to April 17, 2026.
Evertune is the AI marketing platform for Generative Engine Optimization (GEO) that helps brands improve visibility in AI search by analyzing responses at scale and delivering actionable insights. Evertune works with leading brands across all verticals, including Finance, Retail and E-Commerce, Automotive, Pharma, Tech, Travel, Food and Beverage, Entertainment, CPG, and B2B. Founded by early team members of The Trade Desk, Evertune has raised $20M in funding from leading adtech and martech investors. Headquartered in New York City, the company has a growing team of more than 40 employees.
Don't forget to share this post
