©Ralf Hahn via Canva.com
March 20, 2025
Should AI Be Used To Respond to Customer Reviews?
For the second year in a row, BrightLocal’s Local Consumer Review Survey in a blind test found that a majority of consumers prefer responses to online reviews from businesses created by AI rather than a human.
BrightLocal’s annual study of online reviews, in its 15th year, continued to find significant benefits to businesses responding to online customer reviews. Of the just over 1,000 U.S. adults surveyed, 89% were more likely to shop at a business that responds to all reviews, whether positive or negative. About 56% were more hesitant to shop at a business that didn’t respond to reviews.
In its 2024 survey, BrightLocal first introduced a blind test asking its consumer panel to choose a preferred response to a review for a restaurant — one was a real business owner’s response, sourced from Google, and the second was generated using an AI prompt in ChatGPT. Of the respondents, 58% preferred the ChatGPT-generated response versus 42% for the human response.
In the 2025 survey, the blind test explored a response by a vet clinic and saw the same results — the ChatGPT-generated response was preferred to the human response, 58% to 42%, respectively.
Here are the two review responses provided in this year’s survey:
- First response (written by a human): “Denise, we appreciate the confidence you have in Dr. Evans and the entire team at Animal Hospital. We value senior dogs as much as young pups and do everything in our power to keep them all healthy. Thanks so much for your kind remarks and terrific five-star review. It means [so] much to us.”
- Second response (written by AI): “Thank you so much, Denise, for your kind words and for trusting us with the care of your senior pup. Dr. Evans and our entire team truly enjoy being here for you both, and we’re so happy to hear that you feel comfortable with the care we provide. Supporting pets and their families is what we’re all about, and it means the world to us to know we’ve earned your trust. We look forward to many more happy visits with you and your furry companion! Warm regards, Animal Hospital”
BrightLocal said the results contrast with other findings in the report showing that consumers might think reviews written by AI are fake. When asked how they tell if a customer review is fake, 46% of respondents said if it “feels like it was written by AI.”
“Are consumers just less aware of what AI really is than they thought?” BrightLocal speculated in its 2025 study. “As we mused in the 2024 report, it’s hard to say exactly why consumers might prefer this response over the other. Perhaps some consumers felt that the response in option two was more heartfelt, which, when we’re talking about our beloved pets, makes sense! With these findings two years in a row, it does show that AI tools can be a useful aid to help local businesses and marketers respond to their reviews with unique content, at scale.”
A survey of 700 U.S. adults in January 2024 from SEMrush, a company specializing in online visibility management, similarly found that AI-generated content was preferred in blind tests by humans over human-generated content for numerous advertising functions, including social media posts, social media ads, and product descriptions.
The use of AI would be a significant time saver for business owners as well as possibly a better option than outsourcing responses to online customer reviews to a junior employee. Sprout Social in a blog entry noted that AI can provide auto-generated replies to online reviews that reflect the sentiment of the customer’s review and can be tailored to sound “friendlier, shorter, [and] more professional.”
However, surveys show that while people are growing more aware and comfortable with AI, many still have trust issues and want transparency.
A Nielsen survey of 6,000 U.S. adults on AI in media found that 55% felt uncomfortable on websites that rely heavily on AI-generated articles and stories, according to Forbes.
A Yahoo-commissioned poll of over 1,200 U.S. consumers and more than 350 U.S. advertisers from fall 2023 found that while 77% of advertisers view AI positively, only 38% of consumers shared this sentiment. The study also found that 53% of consumers were unfamiliar with AI use in advertisements.
Yet, Yahoo’s study found that AI-generated ads with noticed disclosures significantly boosted brand perceptions and provided a 47% lift in ad appeal, a 73% lift in ad trustworthiness, and a 96% lift in overall trust for the company.
As far as customer reviews, a recent Yelp-commissioned survey of 2,350 Yelp review readers found that a whopping 88% oppose AI-generated reviews on review platforms.
Discussion Questions
Are you generally for or against restaurants or retailers tapping AI in responding to online customer reviews?
Does it make sense that AI-generated responses would be preferred to human ones in most cases?
Poll
Loading …BrainTrust
Joan Treistman
President, The Treistman Group LLC
Nolan Wheeler
Founder and CEO, SYNQ
Georganne Bender
Principal, KIZER & BENDER Speaking
Recent Discussions
More Discussions
Should Under Armour Have Broken Up With Steph Curry?
Under Armour is terminating its relationship with NBA star Steph Curry. Is this a smart move?How Successfully Can Arts and Crafts Retailers Capitalize on a ‘Grandma Hobbies’ Trend?
A so-called ‘grandma hobbies’ trend is apparently gaining traction. How can arts and crafts retailers capitalize…Is the Gen Z ‘Stare’ or Lack of ‘Handshake’ Skills a Real Problem in Retail?
Gen Z has been accused by some of lacking soft skills in retail or service roles…
It depends. I don’t see all that much harm if AI is responding to add context, answer a question, or something similar – provided it can do so appropriately. It’s somewhat impersonal, but then so are stock responses that are often cut and pasted. There will always be occasions where the personal touch is needed so, as usual, AI is not the complete solution.
There are plenty of retail counsels general who want to move cautiously because of AI’s occasional hallucinations. AI for reviews is coming, but perhaps not right away.
I see it for the reasons you state, but where empathy is needed, I don’t think AI can provide that.
As long as customer feedback gets relayed back to management, it makes sense for retailers and restaurants to use AI to handle reviews. Retail teams are already lean due to the labor shortage; why not free up resources by applying AI? The given example shows how far AI has come by expressing empathy and warmth.
Studies like this are generally quite invalid. When studies into empathy are evaluated, it turns out AI is given the advantage because humans are not allowed to use the whole of their humanity – but are forced to reply only as robustly as AI can. In this case, a small business owner has many ways they can reply which re highly effective – for example by contacting the individual. Yet handcuffs are place on human response so it’s no wonder AI is “preferred”.
there is a further error here. The survey company hand picked the AI replies – I expect using tremendous prompt sophistication. In real life, there would be no such sophistication and the business would be left at the mercy of the AI.
Retailers must be very careful about research like this. Perhaps AI is useful as part of their responses. But that takes incredible care and savvy.
An AI system can efficiently handle a large volume of reviews and provide quick, standardized responses, ensuring that each customer feels acknowledged.
Human responses, however, often convey empathy and a personal touch that artificial intelligence may not be able to provide, which is crucial in building a genuine relationship with a customer. The use of artificial intelligence can streamline the process, but a balance between AI and human oversight might provide the best of both worlds.
Artificial intelligence may be able to replicate human empathy more effectively by incorporating natural language processing techniques that understand sentiment. Additionally, integrating machine learning models trained on diverse emotional expressions will enable AI to respond in a way that feels more personalized. The ability of AI to provide emotional responses can be further enhanced by regularly updating these models with feedback from human interaction.
This is an excellent use case for AI. Human responses rely on the skill of the human to make a clear, empathetic and grammatically correct response. Depending on who has response duty that day and how their day is going, the resulting responses will vary.
AI will deliver consistent responses, promptly and use proper grammar.
The key to successful use of AI for responses is a well-crafted prompt and periodic monitoring to keep the responses fresh on point. “Avoid becoming stale and repetitive” can be included in the prompt.
Should AI be used to respond to customer reviews? Absolutely not. Nothing says “we couldn’t care less about what you think” than an automated response created by AI.
Customers are important to every business; they deserve a real response from a real human being.
People that have questions, want good answers. Let’s be honest…AI provides pretty damn good answers for most simpler, repetitive things people often ask. That’s been my experience. And as AI becomes even more capable (better or as good as humans say the biggest brains in the domain in the next 5-10 years), this will be the case. AI in all it’s forms, will do things as good as, or even better than humans in lots of different areas. Ironically, AI will be more “human” than actual humans. This will be great for customers and bad for customer service workers.
AI responses to reviews don’t have constraints as do those by humans who are worried about how long it takes to write a response. Verbosity and flowery language can flow from AI more easily than from a person who manages a restaurant. It’s understandable that customers might prefer the AI versions. However, if the responses become repetitive, they will lose their “authenticity”.
If the response to a review seems AI-generated, the company has failed. It shows a low or no level of authenticity that will rub some customers the wrong way. My take on AI is to be transparent when using it. For positive reviews, acknowledge them with something nice and seemingly personal. AI is great for that. If the review is negative, a generic response is bad, regardless of whether it comes from a human or AI. Let the customer know that they have been heard and offer up a resolution to the complaint. If AI can do that, great. However, usually, a human needs to step in to “save the day.” Regardless, negative reviews need to be dealt with properly. Finally, AI has the ability to not just respond to reviews, but assess all reviews to spot trends (good and bad) that can be helpful for a company/brand. That’s important information that, when used to make changes, can lead to better customer experiences that get even more good reviews.
Looking forward to reading AI-generated reviews being responded to by AI.
Wish we had a laughing emoji here.
I don’t understand why retailers don’t start with replacing scripted CSRs with AI. Scripted CSR’s who barely can be understood because they have such thick accents are crazy-making. And the automated message begging us to accept a link since “most of your questions can be answered by our automated system online…would you like me to text you a link?” is also not good for my blood pressure.
I care less about responses to reviews than I do with responses to my problems.
For or against won’t matter. If there’s one thing Retailers are good at, it’s cutting expenses. AI chatbots will both replace people and allow Retailers to perform more tasks that may not have been possible before, like replying to reviews. To Paula’s point when the bot is not up to the task and a real human with discretion is required – who will be there to turn on the lights and answer?
AI can be a powerful tool when used correctly, as long as the responses remain genuine. If humans work with AI to write responses, they can reduce manual effort while keeping that personal touch.
Anyone that uses Gen AI regularly knows that, if properly trained, both in content and in tone, style and language, that these tools can absolutely respond to customer reviews properly and at scale. There’s nothing overly human about a person with a scripted response so technology should certainly be an option.
AI has its place in managing customer reviews, but it depends. It works well for speed, consistency, and handling volume, but trust isn’t built on automation alone. A well-written AI response may sound polished, but if it lacks genuine connection, it risks feeling transactional, just another automated reply.
The key is knowing when to use AI and when a human touch is essential. AI can handle routine interactions, but moments that require nuance, empathy, or deeper engagement should never be outsourced to a machine. Businesses that get this balance right won’t just respond efficiently, they’ll build real and lasting customer relationships.
Broadly, AI is an excellent tool for retailers and restaurants looking to innovate and streamline their operations. Regarding AI for responding to customer reviews, retailers need to balance AI-optimized operations with providing a personalized customer experience. With this in mind, AI can be a great tool for responding to reviews that don’t require a very personalized response. It may help more customers feel their feedback is seen and acknowledged. Retailers still need to keep a human-in-the-loop approach, leaning on our intuition to understand when certain responses might need a more nuanced and personal touch.