# How Reviews and Mentions Shape AI Recommendations > How genuine reviews, replies, and real mentions on other sites feed AI answers about your business, and which shortcuts break platform rules or federal law. Canonical: https://vassalmarketing.com/information/answer-engine-optimization/reviews-mentions-and-ai-recommendations/ **Maintained by Vassal Marketing** ยท Last updated on September 26, 2026 Reviewed at least quarterly and whenever a platform or rule it cites changes. General information, not legal or professional advice. Check rules that affect your business with a qualified professional. Ask every customer for an honest review, reply to reviews, and earn real mentions from local news, associations, and community groups. AI assistants summarize what customers and other sites say about you, and fake or paid praise can break platform rules and federal law. ## Why this matters AI assistants now summarize your reviews for customers. Google Maps uses Gemini to summarize reviews and other information it finds online, and ChatGPT shows Yelp reviews and ratings for many local searches. BrightLocal's 2026 consumer survey found half of AI users trust AI review summaries, and nearly all sometimes check the reviews themselves. Specific, recent reviews give those summaries something accurate to say. Assistants also lean on what other sites say about you. A 2025 research preprint from the University of Toronto found AI search engines cite third-party sources far more than a brand's own pages. For a local business, that means local news, trade associations, community groups, and review sites. Google warns that seeking inauthentic mentions isn't as helpful as it might seem, and the Federal Trade Commission (FTC) bans fake and undisclosed insider reviews. ## Responsible practices ### Ask every customer, steadily Send the same short, neutral review request to every customer at the same point, such as the day the job is finished. A steady flow keeps reviews recent. Google bans rewards for reviews and asking only happy customers, and Yelp's rules say businesses should not ask for Yelp reviews at all. ### Make it easy to be specific AI summaries draw on review text, so detailed reviews help more than "great service." You can name the job in your request, such as "your water heater replacement," without suggesting what to say about it. Never give customers wording, keywords, or a star rating to use. ### Reply to reviews without private details Replies show customers and assistants that someone is paying attention. One vendor study of large multi-location brands found locations with low ratings and very low reply rates were almost absent from AI recommendations; it may not reflect single-location businesses. Keep replies short and never confirm private details about a customer. ### Earn real mentions Local news coverage, trade association and chamber directories, community sponsorships, "best of" lists, and a Better Business Bureau profile all describe you in someone else's words. In one 2026 survey of 47 local search experts, which reflects opinion rather than measurement, "best of" lists and mentions in news and blogs ranked among the top factors for AI visibility. Earn these by doing newsworthy work and joining groups you actually belong to, not by paying for placement. ### Join community conversations honestly Answering a question on Reddit or Nextdoor can help people, as long as you say you own the business. Do not post as a fake customer, pay others to praise you, or have staff post without saying who they are. The FTC's rule covers fake reviews, undisclosed insider reviews, and fake social media influence. ### Protect privacy in health and legal fields Health, legal, and other licensed businesses face extra rules on what they can say in public. The U.S. Department of Health and Human Services settled with a dental practice that disclosed patient information in replies to Yelp reviews. Have a written reply policy and check it with a qualified adviser. ## Worked example: A quarter of honest reputation work Hypothetical example: a small roofing contractor wants AI assistants to describe its work accurately. Over three months, the owner builds a simple routine. ### Review requests Texts every customer the same message the day a job is finished, naming the job type and linking to Google. ### Replies Replies to each new review within a week, thanking the reviewer and mentioning what they wrote, with no address or job details. ### Local mentions Joins the regional roofing contractors' association and is listed in its member directory. The local paper quotes the owner in a story about storm repair scams. ### Community Answers a Nextdoor question about roof inspections, starting with "I own a roofing company, so take this with that in mind." ### Check Asks Google AI Mode and ChatGPT about roofers in town and reads how the business is summarized. None of this is a trick. It gives assistants accurate, recent material written by customers and other sites. ## Reputation routine checklist ### What you can do yourself 1. Send the same neutral review request to every customer who finished a job this week. 2. Reply to every review from the past month without mentioning private details. 3. List three local groups, associations, or events where your business genuinely belongs. 4. Ask ChatGPT and Google AI Mode to summarize your reviews, and compare with what customers actually wrote. 5. Tell staff not to post reviews or recommendations about the business without saying they work there. ### What to ask your provider or staff to handle 1. Confirm in writing that no review tool or campaign filters customers, offers rewards, or writes reviews. 2. Show you every review reply before it is posted, or agree which kinds they may post without asking. 3. Report any mention, PR, or community campaign to you, including who is paid and how it is disclosed. ## Avoid these mistakes - Buying reviews, or letting a vendor write them with AI - Paying for Reddit posts or "best of" list placements that hide the payment - Sending review links only to customers who seem happy - Posting as a customer or asking relatives to vouch for you - Hiding text on your website meant to steer AI answers AI assistants decide what to show, change often, and can give different answers to different people, so mentions, citations, and recommendations are not guaranteed. ## Questions business owners ask ### Do more reviews mean AI will recommend me? Reviews are one input, not a switch. Assistants summarize what reviews say, so recent, detailed reviews help an assistant describe you accurately. Accurate listings, a clear website, and mentions elsewhere matter too. ### Can I pay for mentions or "best of" list placements? Paid placements must be clearly disclosed, and Google says seeking inauthentic mentions isn't as helpful as it might seem. Fake or undisclosed paid reviews and endorsements can break the FTC's rule and its Endorsement Guides. Earn mentions through real work, memberships, and community involvement instead. ### What if an AI summary of my reviews is wrong? Check whether the reviews themselves support it. If the summary misstates what customers wrote, use the platform's feedback option, such as the thumbs-down on Google's AI answers. If a review breaks the site's rules, report the review. ### Can staff post about the business on Reddit or Nextdoor? Yes, if they clearly say they work for you and are honest. Posting as a customer, or having staff praise the business without saying who they are, can break the FTC's rule and the site's own rules. ## Primary guidance - [Google Maps contribution policy](https://support.google.com/contributionpolicy/answer/16597558) - [Google Search Central: Optimizing your website for generative AI features](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide) - [FTC: Consumer Reviews and Testimonials Rule Q&A](https://www.ftc.gov/business-guidance/resources/consumer-reviews-testimonials-rule-questions-answers) - [HHS: New Vision Dental resolution agreement](https://www.hhs.gov/hipaa/for-professionals/compliance-enforcement/agreements/new-vision/index.html) - [BrightLocal industry survey: consumer trust in AI recommendations (2026)](https://www.brightlocal.com/research/lcrs-ai-trust/) - [Chen et al. research preprint: Generative Engine Optimization (2025)](https://arxiv.org/abs/2509.08919) ## Connected guidance - [Responsible review requests](https://vassalmarketing.com/information/reviews-reputation/responsible-review-requests.md): Sample wording and the rules for asking every customer for an honest review. - [Responding to customer reviews](https://vassalmarketing.com/information/reviews-reputation/responding-to-customer-reviews.md): How to reply in public without sharing private customer details. - [The FTC review rule explained](https://vassalmarketing.com/information/reviews-reputation/ftc-review-rule-explained.md): What the federal rule bans, including fake, paid, and undisclosed insider reviews.