Data Studies
Open research on how AI search engines cite the web. Every report is updated daily from live data across ChatGPT, Perplexity, Gemini, Google AI Mode and AI Overviews.
Each report is broken down across six views — one aggregated and one per AI model.
Which domains AI search engines pull from most often.
Top Cited Domains
The 10 domains most frequently cited in AI search answers, updated daily.
Social platforms in AI answers
Reddit, X, LinkedIn, TikTok, YouTube and more — the social platforms AI search engines lean on.
UGC in AI answers
Reddit, Quora, Wikipedia, Stack Overflow and the forums AI search trusts most.
News publishers in AI answers
Which editorial outlets AI search reaches for when it answers.
Which products and merchants AI recommends in shopping-intent queries.
Aggregate metrics on how AI search engines behave.
Google AI Overviews
Share of queries where Google actually serves an AI Overview on top of the regular SERP.
Citations per answer
How many sources each AI search engine links to on average.
Query fan-out
How many related sub-queries each AI search engine fans out to per answer.
Live tracker of bots that crawl the web.
LLM Pulse runs thousands of prompts against ChatGPT, Perplexity, Gemini, Google AI Mode and AI Overviews each week and captures the full response.
Every link in every answer is parsed, deduplicated and rolled up to the registrable domain. We publish the 28-day rolling aggregate.
A scheduled job rebuilds the rankings every night so every report on this page is never more than 24 hours stale.
The data behind this page
These reports are powered by the same pipeline enterprise brands use inside LLM Pulse — millions of AI answers scraped weekly across ChatGPT, Perplexity, Gemini, Google AI Mode and AI Overviews, normalized to the domain, the page and the brand. Get the private version of this dataset for your own company.