Why Reviews Need Context to Help AI Recommendations
Reviews help AI recommendations most when they say something specific: the service a customer got, the town, how fast you came. Google says review count and ratings feed local ranking, and an AI assistant that searches the web can only repeat details a review actually contains. A bare star rating gives it nothing to say about you.
Two fictional dentists in the same suburb both have a 4.8 average. One has 200 reviews that mostly say "Great!" The other has 90, and a dozen of the recent ones mention Saturday appointments, sedation for anxious patients, and a specific insurance plan. A parent asks an AI assistant for "a dentist near me open Saturdays who's good with nervous kids." The star average is a tie. The text isn't.
Reviews help an AI recommendation when they confirm something specific about you, belong clearly to your business, and are recent enough to describe how you work now. This article covers each of those, the rules for asking, and the situations where reviews mislead.
Do Google reviews help my business get recommended by AI?
Google reviews can help: Google says local results are "mainly based on relevance, distance, and popularity," and review count and rating feed that last factor, while specific reviews explain why a customer chose you. In Google's words, prominence is "based on info like how many websites link to your business and how many reviews you have. More reviews and positive ratings can help your business's local ranking."
Google's ranking page describes Google's own local results. It isn't a published formula for every AI assistant, and no assistant publishes one. What's consistent is that assistants which search the web read the same review pages anyone can see. A review that says "replaced our water heater in Allen the same day" gives them a service, a place and a speed to repeat. "Five stars, great guys" gives them a rating and nothing else.
How can I get more useful reviews without breaking the rules?
Ask every customer, ask them to mention the service they got, reply with specifics, and never offer anything in exchange for a review.
Two sets of rules apply, and they aren't the same:
- Google's policy bans any incentive. Its prohibited content policy says businesses must not "offer incentives – such as payment, discounts, free goods and/or services - in exchange for posting any review," and must not "discourage or prohibit negative reviews, or selectively solicit positive reviews from customers." That last part rules out review gating: asking "were you happy?" first and only sending the review link to people who say yes.
- The FTC's rule is narrower on incentives and backed by penalties. The FTC's final rule, announced Aug 14, 2024, bans fake reviews and testimonials and "prohibits businesses from providing compensation or other incentives conditioned on the writing of consumer reviews expressing a particular sentiment." It's codified at 16 CFR Part 465, took effect Oct 21, 2024, and lets the FTC seek civil penalties against knowing violators.
So a "$10 off for a five-star review" offer breaks both rules, and any reward for a review breaks Google's policy. The simple rule is to offer nothing.
What you can do:
- Use Google's own link. Google says you can ask customers "to visit a Google link or scan a QR code." Put it on the invoice, the follow-up text and the receipt.
- Ask at the right moment, when the job is done and the customer is pleased, not three weeks later.
- Prompt for detail, not sentiment. "If you leave a review, it helps other customers if you mention which service we did and where" is fine. Writing the review for them isn't.
- Reply with specifics. "Thanks for having us out to Frisco for the furnace tune-up" adds your service and town to the page, in your own words.
What should I do about a run of bad reviews?
Fix the cause first, then reply. A cluster of recent complaints about the same thing (missed appointments, surprise fees, a rude technician) can outweigh years of older praise, because it suggests the business has changed.
Reply to each one calmly and specifically, without arguing or sharing private details: say what happened, what you changed, and how to reach someone. Don't ask happy customers to "drown out" the bad reviews in a sudden burst. A burst looks less natural than a steady pace, and asking only satisfied customers is exactly the selective solicitation Google's policy bans. Keep asking everyone, as before.
Mixed reviews are normal. One complaint about parking among eighty good reviews isn't a pattern. Five complaints about billing in a month are.
What if reviews meant for another business show up on my profile?
Flag them on the platform, since reviews only help when they're connected to the right business and duplicate or similarly named listings scramble that. Google says that if you find a review that violates its policies, "you can report it for removal". Reviews about a different business are off-topic and can be reported.
Common versions of this problem:
- Same name, different town. "Summit Plumbing" in Denver gets reviews meant for "Summit Plumbing" in Dallas. Check that your profile shows your website and phone, so customers and systems can tell the two apart.
- Duplicate profiles. An old unverified listing collects reviews that never reach your main profile. Ask the platform to merge or remove the duplicate.
- An old location. After a move, reviews of the old address can stay attached to a listing customers can still find. Mark the old location as moved or closed, rather than leaving both live.
Do reviews for one location count for my other locations?
No. Each branch's reviews describe that branch, so keep them separate and never present one location's rating as the whole company's. A restaurant group with a 4.7 downtown and a 3.9 at the airport has two reputations, not an average.
The same goes for ratings that aren't customer reviews. Employee ratings, app-store ratings for your booking app, and coupon-site badges may all show up when someone searches "[your name] reviews." They measure different things. An employer rating doesn't describe your customer service, and a 4.5 for your app doesn't describe the haircut.
How do reviews and my own website work together?
Reviews are strongest when they confirm something your website already says. If your site says "Saturday appointments" and recent reviews mention Saturday visits, the two back each other up. If reviews keep praising a service your site never mentions, you've found a strength to write up. If reviews say you close at 4 and your listing says 6, one of them is wrong. Find out which and fix it.
Because reviews have to match the business they describe, reviews belong next to your other core facts, which The Business Facts AI Must Agree On covers. For the wider checklist, including listings and crawler access, see How to Show Up When Customers Ask ChatGPT for a Local Recommendation.
The Runexus free quick snapshot includes a small search for independent profiles and reviews tied to your business, so you can see which ones it finds and whether they match your site.