Chatbot Conversations Are Capturing Consumer Intent Before Brands Even Know It
CANNES, France — For decades, that moment of a consumer intent decision lived inside a search query. Now it’s happening inside a conversation with an AI assistant.
When a consumer asks ChatGPT whether a particular skincare product will work for their specific concerns, they’re revealing a bit more than a keyword ever could. This new kind of intent signal, once the exclusive province of search engines and social platforms, is migrating to conversational AI.
“The companies that own that path toward intent and can mine that intent and can understand that intent, whether it’s a conversational environment or whether it’s following the path back and forth from AI to other destinations, that’s where the future lies,” said Debra Aho Williamson, founder and chief analyst at Sonata Insights, during her AI Summit at the Beet.TV Leadership Sessions during Cannes Lions 2026.
A birthday morning revelation
Williamson, who spent 19 years at eMarketer building out the firm’s social media research practice, has turned her analytical lens toward how AI is reshaping consumer behavior. She shared a personal anecdote that illustrates the shift she’s tracking.
Standing in front of her mirror one Saturday morning before her birthday, Williamson noticed she wasn’t happy with how her eyes looked. Shortly after, she saw an Instagram ad featuring Gwyneth Paltrow wearing eye patches for the Goop brand. Rather than clicking through, she took her question to ChatGPT.
“I said:
‘Hey, ChatGPT, tell me about these Goop eye patches. Will they work for me? I have deep-set eyes, I have wrinkles, I have some circles under my eyes.’
“And what ChatGPT said was:
‘No, this is not the product for you. You need this other product that I had never heard of, The Ordinary Caffeine Solution.’
”I didn’t know what that was – never heard of The Ordinary, never thought about using caffeine, but I was intrigued.”
The intent signal black hole
Williamson eventually purchased the recommended products on Amazon. But the transaction data tells only a fraction of the story.
“The retailer at the end does capture the transaction. Amazon got a sale,” Williamson said. “But Amazon and no other media platform other than ChatGPT got any signal, any intent signal of what I was interested in doing.”
What are the implications here? ChatGPT learned that Williamson had a concern, was open to new products, and was ready to buy. That progression of intent, which once would have been captured across multiple search queries and retargeting cookies, happened entirely within a single conversational thread invisible to traditional ad platforms.
AI advertising is scaling, but slowly
The shift Williamson describes is beginning to attract real money. US AI ad spending will more than double over five years to $68.25 billion in 2030, according to Emarketer’s 2026 forecast. However, the research firm notes that in 2026, more than 80% of AI advertising will appear adjacent to AI content rather than inside chatbot conversations themselves.
Chatbot-native ad spending remains a small but fast-growing category. Emarketer projects AI chatbot ad spending will rise 1,641% in 2026, but only to $0.96 billion. The larger near-term opportunity sits in AI search-adjacent advertising, forecast to grow 152% to $26.42 billion this year.
Following the consumer, following the money
Williamson, who launched her AI Ad Economy Substack earlier this year, argues that marketers need to pay attention to where consumer behavior is heading, not just where budgets currently sit.
“I also want to encourage you to just think about the consumer,” she said. “Think about how consumer behavior is changing. Think about how your advertising is reaching them and what signals you are able to adapt to.”
She acknowledged that advertising within AI platforms like ChatGPT isn’t right for every brand. But she emphasized that the underlying behavioral shift is real and durable.
“When consumer behavior changes and technology changes, marketing changes too,” Williamson said. “And following the consumer is going to, I think, help you to follow where the money is going as well.”
CANNES, France — For decades, that moment of a consumer intent decision lived inside a search query. Now it’s happening inside a conversation with an AI assistant.
When a consumer asks ChatGPT whether a particular skincare product will work for their specific concerns, they’re revealing a bit more than a keyword ever could. This new kind of intent signal, once the exclusive province of search engines and social platforms, is migrating to conversational AI.
“The companies that own that path toward intent and can mine that intent and can understand that intent, whether it’s a conversational environment or whether it’s following the path back and forth from AI to other destinations, that’s where the future lies,” said Debra Aho Williamson, founder and chief analyst at Sonata Insights, during her AI Summit at the Beet.TV Leadership Sessions during Cannes Lions 2026.
A birthday morning revelation
Williamson, who spent 19 years at eMarketer building out the firm’s social media research practice, has turned her analytical lens toward how AI is reshaping consumer behavior. She shared a personal anecdote that illustrates the shift she’s tracking.
Standing in front of her mirror one Saturday morning before her birthday, Williamson noticed she wasn’t happy with how her eyes looked. Shortly after, she saw an Instagram ad featuring Gwyneth Paltrow wearing eye patches for the Goop brand. Rather than clicking through, she took her question to ChatGPT.
“I said:
‘Hey, ChatGPT, tell me about these Goop eye patches. Will they work for me? I have deep-set eyes, I have wrinkles, I have some circles under my eyes.’
“And what ChatGPT said was:
‘No, this is not the product for you. You need this other product that I had never heard of, The Ordinary Caffeine Solution.’
”I didn’t know what that was – never heard of The Ordinary, never thought about using caffeine, but I was intrigued.”
The intent signal black hole
Williamson eventually purchased the recommended products on Amazon. But the transaction data tells only a fraction of the story.
“The retailer at the end does capture the transaction. Amazon got a sale,” Williamson said. “But Amazon and no other media platform other than ChatGPT got any signal, any intent signal of what I was interested in doing.”
What are the implications here? ChatGPT learned that Williamson had a concern, was open to new products, and was ready to buy. That progression of intent, which once would have been captured across multiple search queries and retargeting cookies, happened entirely within a single conversational thread invisible to traditional ad platforms.
AI advertising is scaling, but slowly
The shift Williamson describes is beginning to attract real money. US AI ad spending will more than double over five years to $68.25 billion in 2030, according to Emarketer’s 2026 forecast. However, the research firm notes that in 2026, more than 80% of AI advertising will appear adjacent to AI content rather than inside chatbot conversations themselves.
Chatbot-native ad spending remains a small but fast-growing category. Emarketer projects AI chatbot ad spending will rise 1,641% in 2026, but only to $0.96 billion. The larger near-term opportunity sits in AI search-adjacent advertising, forecast to grow 152% to $26.42 billion this year.
Following the consumer, following the money
Williamson, who launched her AI Ad Economy Substack earlier this year, argues that marketers need to pay attention to where consumer behavior is heading, not just where budgets currently sit.
“I also want to encourage you to just think about the consumer,” she said. “Think about how consumer behavior is changing. Think about how your advertising is reaching them and what signals you are able to adapt to.”
She acknowledged that advertising within AI platforms like ChatGPT isn’t right for every brand. But she emphasized that the underlying behavioral shift is real and durable.
“When consumer behavior changes and technology changes, marketing changes too,” Williamson said. “And following the consumer is going to, I think, help you to follow where the money is going as well.”
Series Videos
