Marketers have been focused on getting brands into AI answers. New survey data from Semrush suggests the bigger risk is getting talked out of the sale.
In a survey of 2,338 U.S. adults, 81% of consumers who use chatbots while shopping online say they have decided against a purchase based on AI guidance. Even among AI users broadly, more than half have walked away from a product after what a chatbot told them.
AI works as a dealmaker and a dealbreaker
AI recommendations cut both ways. Roughly 58% of AI users have purchased something a chatbot recommended, but the same share have decided against a purchase because of AI-provided information. Among people who use AI multiple times a week, more than 80% have bought an AI-recommended product.
Negative or mixed reviews can make the downside worse. More than 74% of all consumers said they would be at least somewhat less likely to buy a product if a chatbot surfaced mixed or negative reviews. That jumps to more than 85% among AI users.
The study also finds that AI is a discovery channel: nearly 59% of AI users have found a new brand or product through a chatbot recommendation, and about 71% believe chatbots can recommend the best brands and products.
Shopping research is getting more fragmented
This is not a story of AI replacing Google. Only 29% of all consumers use AI chatbots for product research, while among weekly AI users the figure is 55%. For that weekly group, chatbots already outrank friends and family, social media, retail websites, and YouTube for product research, though search engines and reviews still lead.
Meanwhile, 65% of AI users say chatbots have replaced at least some of their product-related Google searches, but almost half said their overall Google use for purchase research has not changed in the past year.
Shoppers are adding AI to their path to purchase rather than swapping one channel for another. A single purchase decision can bounce among Google, reviews, retail sites, and AI, creating more moments where a brand can win or lose consideration.
Chasing visibility is not enough
Buying ads inside chatbots may generate visibility, but it does not control the answers that come five minutes later when a shopper asks whether a product is worth the price or how it compares with a competitor.
Trust is another complication. About 42% of consumers dislike chatbot ads, while 20% like them and 38% are neutral. Among those who dislike ads, 66% say ads make them question the integrity of AI responses.
What marketers should do
Instead of treating AI only as a paid placement opportunity, brand teams need to manage their presence across the sources chatbots are likely to draw from. That includes:
- Monitoring how AI tools describe the brand and its category in purchase-related queries.
- Addressing the specific review themes and complaints that surface in AI summaries.
- Feeding accurate, up-to-date product information into visible sources such as brand pages, retailer listings, and third-party reviews.
- Testing how product comparisons play out when a chatbot is asked to recommend between your brand and a competitor.
The survey makes the mandate clear: consumers are not only asking AI what to buy, they are also asking whether they should buy from you at all. Brands that ignore the answer risk losing consideration before a shopper ever reaches the cart.
Source: MarTech
