Copy the formatted transcript to paste into ChatGPT or Claude for analysis
So I have some AI search and SEO news for you.
ChakGBT just released a brand new paper
showing the usage inside of their platform
and showing exactly how users are using ChakGBT
on a daily basis.
In fact, this is data from over 700 million
weekly active users.
So there's a lot of little insights we can pull from this
to help us drive more brand awareness inside of ChakGBT
and ultimately do AI search optimization.
So a couple of things here.
First is 49% of the messages within ChakGBT
are what are considered asking intent, okay?
So this is gonna be sometimes
very informational type of questions,
AI-generated overview
Nathan Gotch analyzes OpenAI's ChatGPT usage report covering 700 million weekly active users, identifying three intent buckets: (1) Asking (49%)—informational/commercial questions like "who are the best SEO tools for agencies rank by price" where retrieval is used, making these trackable opportunities; (2) Doing (40%)—execution tasks like "set up seat-based billing with trials, give me 3-5 options with pricing and docs" representing sales journey compression where entire buyer journeys happen in one 30-minute chat window; (3) Expressing (11%)—role-play scenarios with zero brand visibility potential. Critical insight: dangerous informational queries exist where ChatGPT answers from core training without retrieval, requiring no citations. The framework emphasizes tracking "asking" queries via Google Search Console regex patterns and understanding that most conversions won't be direct ChatGPT referrals but rather branded searches driven by AI-generated visibility. Gotch argues 95% of ChatGPT users also use Google, positioning the platforms as complementary rather than competitive, with ChatGPT functioning as a marketing channel driving branded search intent.
49% of ChatGPT messages are "asking intent" queries, but "informational queries are very dangerous" because if ChatGPT knows the answer from core training, "it is not going to use retrieval"—requiring focus on fresh queries ChatGPT "cannot respond to in any accurate way without having to go and retrieve answers."
"Who are the best [product category] for [use case] rank by criteria" and "what are the best alternatives to [brand]" are high-value query formats that "surface brands" and require retrieval, making them trackable optimization targets.
"Sales journey compression" enables entire buyer journeys in "one chat window" taking as little as 30 minutes from problem awareness to purchase, compared to the traditional "erratic" multi-platform journey.
The buyer journey in ChatGPT follows problem awareness → solution exploration → specific goals → roadblock → product recommendations, with "money made" only at the final stage when ChatGPT recommends solutions.
🚀 Get 10x better SEO & AI search results: https://nathangotch.com/