April 20, 2026
stokkete/Depositphotos.com
Will Shoppers Ever Stop Second-Guessing AI Brand Recommendations?
A recent Idea Grove study unearthed an interesting and significant statistic: When consumers are presented with an unknown brand following an AI recommendation, nearly all of them (98%) pivoted to verify that recommendation from other trusted sources.
The study, “How Consumers Verify AI-Recommended Brands,” outlined that despite massive effort, energy, and spend being allocated to AI ranking and optimization coming from brands and retailers, the average shopper is still highly skeptical of the results generated by artificial intelligence models.
“The other 98% go looking for something more: reviews, search rankings, press coverage, a website that holds up. The AI recommendation opens the door. What’s on the other side of it determines whether anyone walks through,” the study authors reported.
Other interesting findings pulled from the report:
- Customer reviews are the No. 1 trust signal after AI recommendations: Respondents ranked customer reviews in the top slot (78%) regarding trust signals following an AI brand or product recommendation, even though these can also be subject to fraud or misrepresentation. Next were Google rankings (71%), business longevity (69%), and press coverage (58%).
- Nearly half of U.S. shoppers unaware that AI recommendations are often influenced by agencies: About 48% of those polled did not know that hired agencies were frequently paid to influence AI results, with that stat increasing to 65% among older Americans.
- A generational trust and usage divide exists: While over two-thirds of Gen Zers use ChatGPT for brand research (compared to just 30% of baby boomers), when it comes to trust, younger shoppers are more likely to trust AI recommendations (43% of zoomers and 39% of millennials, versus 18% of boomers).
Skepticism Over AI Brand, Product Recommendations Remains High, But Shows Signs of Abating
Overall skepticism was quite high across the board, regardless of demographic concerns, although that trust is growing, particularly with younger consumers — a third (32%) of respondents indicate that they trust AI brand recommendations more than they had a year prior, while just 16% indicated less trust versus last year. A majority (52%) were not persuaded in one direction or the other.
The largest bloc? Described as pragmatic skeptics (40%) by Idea Grove, these shoppers find AI useful but remain skeptical. Next up: those more skeptical, who believe that certain brands may have “gamed” the algorithm to present themselves in a more positive light (27%). A full 19% said they do not trust AI recommendations whatsoever, while the smallest cohort (15%) placed their faith in AI recommendations as being the best options.
“AI is changing where people start their brand research, but it hasn’t changed what convinces them to buy,” said Scott Baradell, founder and CEO of Idea Grove.
“Consumers treat an AI recommendation as a starting point, then fall back on the same signals that have always built trust — reviews, press coverage, search rankings, a credible web presence. What’s striking is that AI systems themselves were trained on those same signals. The evidence that makes a brand credible to a careful human buyer is the same evidence that makes it credible to the machines. A recommendation opens the door. What brands have built before that moment determines whether anyone walks through it,” he added.
BrainTrust
Tanya Thorson
Revenue & Customer Growth Leader, StrategiX Marketing
Pamela Kaplan
Principal, PK Consulting
Doug Garnett
President, Protonik
Recent Discussions
Discussion Questions
Will shoppers ever stop second-guessing AI recommendations? What will be the turning point, or necessary missing element, if so?
How can brands or retailers best protect themselves against rising backlash against AI partnerships, while still participating in the ecosystem? Is mitigation even necessary, or will adoption solve this problem over time?
Poll
Loading …
The use of AI is growing rapidly. However, this does not signal that AI is trusted: there is a huge amount of skepticism around AI output. From our Pacsun research, even among youth consumers – who are very AI savvy – only 21.3% fully trust AI responses. The lesson for brands is that AI is an additional channel not a replacement channel – so brands need to ensure that other channels continue to pull their weight in terms of supporting consumer decision making. And this pattern is not likely to change in the future because consumers, by nature, use multiple touchpoints when shopping. Of course, it raises an interesting consideration because if trust isn’t present and doesn’t build, it puts a rather big hole in all of these forecasts saying that AI doing our shopping for us will predominate.
Trust must be earned. In no way has AI earned the trust of the most skilled practitioners of AI — much less consumers. That said, this behavior is the same everywhere. When ANYONE makes a brand recommendation I expect 98% of people will go out to verify it on their own — or at least to learn more about what they might be getting into.
Buyer beware is eternal advice and no company should believe it can take action which makes consumer wariness unnecessary. It can be reduced or helped with broader brand trust — but it will never be eliminated.
Even if this thing-which-really-hasn’t-had-a-chance-to-start-yet-so-how could-it-stop? exists, how, exactly, is it a problem?
Over time, and as AI technology becomes more prevalent and more refined, users will question its recommendations less and go along with its recommendations more. It will take more time. By then, the discovery of a new brand recommendation will be like a treasure hunt.
“ more refined” by whom? good actors or bad/competitive/criminal actors?
Is trust inherently linear?
AI can be quite helpful and efficient…but, for now, nothing beats customer feedback. I’ll always want to hear from a person who purchased the product.
Like Ronald Regan said about negotiations with the Soviet Union: “Trust, but verify,” and that sums up this study and where consumers are at today. It’s hard to say how long it will take but eventually trust will build and AI verification will wane. But for now, there’s so much evidence to show that AI messes up, the only prudent thing for any consumer to do it to verify AI results from secondary sources. That said, starting with AI search first is the better way to go, in my personal experience.
I think the better question is what have we done to earn the trust of our consumers? Phrases like “caveat emptor” and “trust but verify” are prevalent because the profit motive often creates a gap between the interests of the seller and the buyer.
I agree with all my colleagues except for Bob. Sorry, Bob. I’m not convinced that it will ever change because the moment it does our temptation to use a trusted resource to skew recommendations will be too great. Or has that happened already?
Who’s feeding the AI? Do you trust them? I don’t, especially knowing that brands pay to influence AI.
AI can point me in the right direction, but I will trust reviews left by customers over those machine-generated every time.
AI recommendations will keep growing, but shoppers are not outsourcing discernment. They are outsourcing the first pass. That is a very different thing.
What this study really shows is that AI has become a discovery layer, not a trust layer. The recommendation may narrow the field, but the shopper still goes looking for proof. That makes the post-AI moment more important, not less. Reviews, search presence, site experience, product detail, and brand consistency now carry even more weight because they have to validate what the machine surfaced.
For brands and retailers, the opportunity is not just to rank. It is to reinforce. If AI is the introduction, the rest of the ecosystem has to close the confidence gap fast. The brands that win will be the ones that make verification easy, credible, and consistent across every touchpoint.
AI can speed up consideration.It cannot shortcut trust.
Discovery is faster. Proof matters more.
People tend to second guess anything that introduces a new brand, whether it comes from search, social, or AI. AI speeds up discovery, but people will still look for validation through reviews and other signals before making a decision.
It’s simple: The secret to successful AI suggestions is relevance and moderation (as in not too much). That will begin building trust, but it may still take a number of VERY relevant suggestions before the customer believes that AI is thinking in their best interest. According to my annual research (https://www.CustomerExperienceResearch.com), the top trusted sources for product recommendations go to family, then reviews/ratings on websites (like Amazon, Walmart, etc.), followed by friends.
AI is great for discovery and comparison, but it’s not trusted today. When it comes to trust, I don’t see a world where there is anything better than customer reviews. What might change is not customers verifying, but how much effort verification takes.
Shoppers will not fully stop second guessing AI recommendations because trust is not built at the moment of recommendation, it is validated through consistency across signals. AI is simply compressing the discovery phase, not replacing human judgment. The turning point will come when AI outputs are explainable, transparent, and consistently aligned with real world signals like reviews, availability, and experience.
From a personalization standpoint, your point is right. Strong data inputs drive strong outputs. Retailers who unify product data, behavioral signals, and credibility markers will see AI act less like a black box and more like a trusted advisor, reducing the need for constant verification over time.
Consumers simply do not trust advertising messages. This isn’t new. And after years of placed results in various search engines and by online retailers, the issue for consumers is even worse; most can smell the hand of marketing a mile away. AI is perceived as an extension of the same tech that serves up the ads they’ve learned to be skeptical about today.
In the words of Frank Zappa — “You will do as you are told, until the rights to you are sold.” Today’s digital‑native generations are extremely savvy and extremely likely to want to verify any messages, from any digital source, when researching a product.
The question really becomes how much do you trust any source? SEO impacts search results, reviews can be gamed, and AI results can be influenced. Seems like the best practice when shopping would be to consider multiple sources, realizing that all of them have a bias.
Customers are unlikely to stop second-guessing AI recommendations, and that is actually expected. AI is changing where customers start their search, but it is not replacing how they make decisions. Most customers will still take a second step to validate what they see, whether that is reviews, search results, or checking the brand itself.
The shift here is important. AI can bring a brand into consideration, but it cannot build trust on its own. That still comes from consistent product information, reliable reviews, and a credible presence across channels. Retailers and brands should focus less on trying to control the recommendation and more on what the customer sees after it. That is what ultimately drives the purchase.
Donn’t I need to develop trust on at least 2 different levels? First, in my own skill at writing prompts that create the right kind of search assignment…??? And second, I will have to experience enough solid product outcomes that I can truly believe AI is performing to my desired outcomes. In the beginning, if I don’t experience solid outcomes, how will I know if my prompt was the problem or if AI just took an odd detour…???
The real issue isn’t consumer skepticism; it’s a structural misalignment of incentives. The 48% of shoppers who are unaware that AI recommendations can be influenced by paid agencies aren’t uninformed; they’re unprotected. Platform-controlled agents, trained to optimize for platform revenue, have every reason to serve curated results dressed as neutral ones. Consumer skepticism is rational precisely because trust signals and profit signals run on the same rails.
Second-guessing won’t fade when AI gets better. It will fade when shoppers can reliably distinguish agents working for them from agents working on them. That’s not a technology problem. It’s a transparency and accountability problem the industry has yet to seriously confront.
“It’s a transparency and accountability problem the industry has yet to seriously confront.”
And one that doesn’t really have an endpoint.
Shoppers second-guessing AI recommendations is actually a healthy sign — it means trust hasn’t been blindly outsourced yet. AI today is very effective at narrowing options, but not at earning conviction. That still comes from external validation like reviews, prior experience, and brand familiarity.
What’s happening is a split in the journey. AI is accelerating discovery and comparison, but the final decision still depends on trust signals outside the AI interface. Until AI can consistently explain why something is recommended — beyond just relevance — users will keep validating elsewhere.
So the gap isn’t just accuracy, it’s confidence and transparency. AI can suggest, but trust still needs to be proven.
“Shoppers second-guessing AI recommendations is actually a healthy sign — it means trust hasn’t been blindly outsourced yet.”
I like your comment. It is encouraging. I fear that AI is destroying critical thinking.
AI is still in its infancy and it shows. Poorly positioned statements and products, key mispellings, verbal garble that is definitively out of place, and no coherence to most AI responses only continue to reinforce the lack of trust in AI postings and usage. This drives consumers to continue to use other consumer reviews, commentary, etc. in place of increasing their trust of AI.
More Discussions
Are ‘Better-For-You’ Product Tweaks Enough To Anchor a Kraft Heinz Turnaround?
Can Kraft Heinz continue its upward trajectory by tweaking its legacy lineup to match today’s tastes?What Could a RadioShack Retail Comeback Look Like?
RadioShack appears to be mounting some degree of growth in B2B and DTC.Will Increased AI Security Halt the Selective Rollback of Self-Checkouts?
Can computer-driven, AI-based security improvements help to lessen self-checkout rollbacks?
The use of AI is growing rapidly. However, this does not signal that AI is trusted: there is a huge amount of skepticism around AI output. From our Pacsun research, even among youth consumers – who are very AI savvy – only 21.3% fully trust AI responses. The lesson for brands is that AI is an additional channel not a replacement channel – so brands need to ensure that other channels continue to pull their weight in terms of supporting consumer decision making. And this pattern is not likely to change in the future because consumers, by nature, use multiple touchpoints when shopping. Of course, it raises an interesting consideration because if trust isn’t present and doesn’t build, it puts a rather big hole in all of these forecasts saying that AI doing our shopping for us will predominate.
Trust must be earned. In no way has AI earned the trust of the most skilled practitioners of AI — much less consumers. That said, this behavior is the same everywhere. When ANYONE makes a brand recommendation I expect 98% of people will go out to verify it on their own — or at least to learn more about what they might be getting into.
Buyer beware is eternal advice and no company should believe it can take action which makes consumer wariness unnecessary. It can be reduced or helped with broader brand trust — but it will never be eliminated.
Even if this thing-which-really-hasn’t-had-a-chance-to-start-yet-so-how could-it-stop? exists, how, exactly, is it a problem?
Over time, and as AI technology becomes more prevalent and more refined, users will question its recommendations less and go along with its recommendations more. It will take more time. By then, the discovery of a new brand recommendation will be like a treasure hunt.
“ more refined” by whom? good actors or bad/competitive/criminal actors?
Is trust inherently linear?
AI can be quite helpful and efficient…but, for now, nothing beats customer feedback. I’ll always want to hear from a person who purchased the product.
Like Ronald Regan said about negotiations with the Soviet Union: “Trust, but verify,” and that sums up this study and where consumers are at today. It’s hard to say how long it will take but eventually trust will build and AI verification will wane. But for now, there’s so much evidence to show that AI messes up, the only prudent thing for any consumer to do it to verify AI results from secondary sources. That said, starting with AI search first is the better way to go, in my personal experience.
I think the better question is what have we done to earn the trust of our consumers? Phrases like “caveat emptor” and “trust but verify” are prevalent because the profit motive often creates a gap between the interests of the seller and the buyer.
I agree with all my colleagues except for Bob. Sorry, Bob. I’m not convinced that it will ever change because the moment it does our temptation to use a trusted resource to skew recommendations will be too great. Or has that happened already?
Who’s feeding the AI? Do you trust them? I don’t, especially knowing that brands pay to influence AI.
AI can point me in the right direction, but I will trust reviews left by customers over those machine-generated every time.
AI recommendations will keep growing, but shoppers are not outsourcing discernment. They are outsourcing the first pass. That is a very different thing.
What this study really shows is that AI has become a discovery layer, not a trust layer. The recommendation may narrow the field, but the shopper still goes looking for proof. That makes the post-AI moment more important, not less. Reviews, search presence, site experience, product detail, and brand consistency now carry even more weight because they have to validate what the machine surfaced.
For brands and retailers, the opportunity is not just to rank. It is to reinforce. If AI is the introduction, the rest of the ecosystem has to close the confidence gap fast. The brands that win will be the ones that make verification easy, credible, and consistent across every touchpoint.
AI can speed up consideration.It cannot shortcut trust.
Discovery is faster. Proof matters more.
People tend to second guess anything that introduces a new brand, whether it comes from search, social, or AI. AI speeds up discovery, but people will still look for validation through reviews and other signals before making a decision.
It’s simple: The secret to successful AI suggestions is relevance and moderation (as in not too much). That will begin building trust, but it may still take a number of VERY relevant suggestions before the customer believes that AI is thinking in their best interest. According to my annual research (https://www.CustomerExperienceResearch.com), the top trusted sources for product recommendations go to family, then reviews/ratings on websites (like Amazon, Walmart, etc.), followed by friends.
AI is great for discovery and comparison, but it’s not trusted today. When it comes to trust, I don’t see a world where there is anything better than customer reviews. What might change is not customers verifying, but how much effort verification takes.
Shoppers will not fully stop second guessing AI recommendations because trust is not built at the moment of recommendation, it is validated through consistency across signals. AI is simply compressing the discovery phase, not replacing human judgment. The turning point will come when AI outputs are explainable, transparent, and consistently aligned with real world signals like reviews, availability, and experience.
From a personalization standpoint, your point is right. Strong data inputs drive strong outputs. Retailers who unify product data, behavioral signals, and credibility markers will see AI act less like a black box and more like a trusted advisor, reducing the need for constant verification over time.
Consumers simply do not trust advertising messages. This isn’t new. And after years of placed results in various search engines and by online retailers, the issue for consumers is even worse; most can smell the hand of marketing a mile away. AI is perceived as an extension of the same tech that serves up the ads they’ve learned to be skeptical about today.
In the words of Frank Zappa — “You will do as you are told, until the rights to you are sold.” Today’s digital‑native generations are extremely savvy and extremely likely to want to verify any messages, from any digital source, when researching a product.
The question really becomes how much do you trust any source? SEO impacts search results, reviews can be gamed, and AI results can be influenced. Seems like the best practice when shopping would be to consider multiple sources, realizing that all of them have a bias.
Customers are unlikely to stop second-guessing AI recommendations, and that is actually expected. AI is changing where customers start their search, but it is not replacing how they make decisions. Most customers will still take a second step to validate what they see, whether that is reviews, search results, or checking the brand itself.
The shift here is important. AI can bring a brand into consideration, but it cannot build trust on its own. That still comes from consistent product information, reliable reviews, and a credible presence across channels. Retailers and brands should focus less on trying to control the recommendation and more on what the customer sees after it. That is what ultimately drives the purchase.
Donn’t I need to develop trust on at least 2 different levels? First, in my own skill at writing prompts that create the right kind of search assignment…??? And second, I will have to experience enough solid product outcomes that I can truly believe AI is performing to my desired outcomes. In the beginning, if I don’t experience solid outcomes, how will I know if my prompt was the problem or if AI just took an odd detour…???
The real issue isn’t consumer skepticism; it’s a structural misalignment of incentives. The 48% of shoppers who are unaware that AI recommendations can be influenced by paid agencies aren’t uninformed; they’re unprotected. Platform-controlled agents, trained to optimize for platform revenue, have every reason to serve curated results dressed as neutral ones. Consumer skepticism is rational precisely because trust signals and profit signals run on the same rails.
Second-guessing won’t fade when AI gets better. It will fade when shoppers can reliably distinguish agents working for them from agents working on them. That’s not a technology problem. It’s a transparency and accountability problem the industry has yet to seriously confront.
“It’s a transparency and accountability problem the industry has yet to seriously confront.”
And one that doesn’t really have an endpoint.
Shoppers second-guessing AI recommendations is actually a healthy sign — it means trust hasn’t been blindly outsourced yet. AI today is very effective at narrowing options, but not at earning conviction. That still comes from external validation like reviews, prior experience, and brand familiarity.
What’s happening is a split in the journey. AI is accelerating discovery and comparison, but the final decision still depends on trust signals outside the AI interface. Until AI can consistently explain why something is recommended — beyond just relevance — users will keep validating elsewhere.
So the gap isn’t just accuracy, it’s confidence and transparency. AI can suggest, but trust still needs to be proven.
“Shoppers second-guessing AI recommendations is actually a healthy sign — it means trust hasn’t been blindly outsourced yet.”
I like your comment. It is encouraging. I fear that AI is destroying critical thinking.
AI is still in its infancy and it shows. Poorly positioned statements and products, key mispellings, verbal garble that is definitively out of place, and no coherence to most AI responses only continue to reinforce the lack of trust in AI postings and usage. This drives consumers to continue to use other consumer reviews, commentary, etc. in place of increasing their trust of AI.