Published 2026-09-23 · Last updated 2026-09-23
Yes, reviews still matter, and they matter more than they used to, because the thing doing the recommending changed. For years a review was something one person read before deciding to call you. Now an AI assistant reads those same public reviews, at a scale no person could, and uses them to decide whose name to say. BrightLocal's Local Consumer Review Survey 2026 found that about 42% of consumers now trust a recommendation from an AI platform as much as they trust a written review. So the reviews you earn are no longer only social proof for people. They are the raw material a machine uses to pick a business, and the machine rewards the genuine ones while it discounts the fakes.
Do reviews still matter now that AI does the recommending?
They matter more, because more customers are asking the machine in the first place. BrightLocal's 2026 survey found the share of consumers using AI to find local business recommendations climbed from 6% in 2025 to 45% this year, which makes AI the third most common way people find a local business, behind Google and Facebook and ahead of everything else. When a homeowner asks an assistant who the best plumber near them is, the answer is built partly from what your reviews say, how many you have, and how recent they are. Your reviews are not a page you keep for the occasional human visitor anymore. They are feedstock for the system that names a business or sends that customer to a competitor.
How does an AI decide which business to name?
It reads your public signals at scale and checks whether they agree with each other. It looks at how many reviews you have, your star average, how recent the reviews are, and what the recent ones actually describe. Then it corroborates. It checks whether your name, phone number, services, and the towns you cover say the same thing on your website, your Google Business Profile, and the other places it can read. Consistency is the underlying signal. A business whose reviews are recent and specific, and whose facts line up everywhere, is easy for the assistant to trust and easy to name. A business that contradicts itself is easy to skip, because the machine has plenty of others to choose from.
What happens to fake or paid-for reviews?
They are exactly what these systems are built to catch and discount, and two separate forces are aimed at them. The first is the law. The FTC's Consumer Reviews and Testimonials Rule, announced in 2024, bans fake and AI-generated reviews, reviews written by company insiders without disclosure, and reviews bought with compensation or incentives conditioned on a particular sentiment. It carries civil penalties. The second is the platforms. Google reported that in 2025 it blocked or removed more than 292 million policy-violating reviews and took down over 13 million fake Business Profiles, and it now pauses new reviews and shows customers a public banner when it detects a sudden spike of spam on a profile. The same statistical muscle that reads reviews to recommend you is the muscle that spots the ones that were manufactured.
That is why the shortcut is a trap. A wall of five-star reviews that all arrived in the same week, worded like they came from the same hand, is the pattern these systems are tuned to distrust. Bought reputation is not an asset you own. It is a liability a platform can wipe, and announce with a banner, on any day it decides to look.
Which reviews help you, and which ones get discounted?
The line is not about star count. It is about whether the review reads like a real customer describing a real job. Here is the difference in plain terms.
| Signal | Reviews that help the AI recommend you | Reviews it discounts or removes |
|---|---|---|
| Who wrote it | A real customer describing an actual job you did | A stranger, a bot, or the owner posing as a customer |
| Timing | A steady trickle as jobs finish, week after week | A sudden burst that trips a spam spike |
| What it says | Specific: the service, the town, how it turned out | Vague praise that could be about any business anywhere |
| The trade | Given freely after good work | Bought, discounted, or written in exchange for a reward |
| Your reply | A calm, specific answer left on the record | Silence, or the same copy and paste under every one |
Both columns are made of the same star ratings. Only the left one compounds. The genuine reviews build a reputation the assistant can read, trust, and keep naming. The manufactured ones build a risk that sits there until a filter finds it.
How much do customers actually trust what the AI says?
Enough to change who gets the call. In BrightLocal's 2026 survey, the 42% who trust an AI platform's recommendation as much as a written review sit alongside a wider tilt: more consumers said they trust AI for local recommendations, around 40%, than said they distrust it, around 32%. The survey also found that 82% of consumers read the AI's review summary, and 23% would decide on that summary alone. That summary is written from your reviews, so a steady flow of recent, detailed, honest ones shapes the sentence the customer reads. Most people still double-check before they act, which is the whole reason the underlying reviews have to hold up when the customer clicks through to read them.
So what should a business actually do about it?
Earn genuine reviews, keep them current, and make sure the machine can read them. None of that is a trick, and all of it is the ordinary reputation work that already wins customers. In practice it looks like this.
- Ask every finished job for an honest review, and never attach a reward to a positive one. A simple email request works, and it keeps you on the right side of the FTC rule.
- Keep a steady flow instead of a one-time push, so recency and a natural pace read as real to both people and the AI.
- Reply to every review on the record, calmly and specifically, because the next reader and the assistant both see that someone is home.
- Make your name, phone, services, and service area match on your website, your Google profile, and every listing, so the facts the AI checks all agree.
Does this replace Google reviews and local SEO?
No. It sits on top of what already works. The local map results still send most trade jobs, your reviews still feed the answers the AI writes, and a clear, fast website still turns a visit into a call. The AI is simply a new reader of the same reputation you have always been building. We build and maintain that reputation for both readers at once: the person who has to trust you in five seconds, and the assistant deciding whether to name you at all. At Optimus we run this as local SEO and reputation work, earning genuine reviews from real customers and surfacing them so Google and the AI can both read and trust them.
Common questions
Can I just buy reviews to get recommended by AI?
No. The FTC's 2024 Consumer Reviews and Testimonials Rule bans fake, AI-generated, insider, and incentivized reviews, and platforms remove them at scale, with Google reporting more than 292 million policy-violating reviews removed in 2025. Bought reviews are a liability the systems are built to discount, not a shortcut to a recommendation.
Is it against the rules to offer a discount in exchange for a review?
Offering compensation or a reward conditioned on a review that expresses a particular sentiment is exactly what the FTC rule prohibits. You are free to ask every customer for an honest review, positive or not. You cannot pay for a good one.
How many reviews do I need before the AI notices me?
There is no magic number, and nobody outside the platforms can promise one. Recency and consistency matter as much as count. A steady flow of recent, specific reviews that match your profile everywhere is what these systems reward, more than a big pile that arrived all at once.
Do a few bad reviews hurt me with the AI?
A handful of honest critical reviews, each with a calm and specific reply, reads as real and human. A wall of nothing but five stars posted in a burst is what looks manufactured. Reply on the record, fix what a fair complaint points to, and keep earning genuine reviews.
Does this replace my Google Business Profile?
No. Your profile and its reviews are a large part of what the AI reads, so keeping it accurate, current, and backed by recent reviews pays off twice, in Google's map results and in the AI's answer at the same time.