Your Five-Star Reviews Have a Shelf Life Now
A five-star page whose newest review is from last winter looks like a business that used to be good. Here is the 30-second test and the weekly fix.
Co-Founder, Vyzz
Reviews go stale
Reviews have a shelf life now. A wall of five-star reviews that stopped coming in last winter still looks great at a glance. But to an AI assistant picking a business to recommend this week, and to the customer who clicks through to check, it reads like a place that used to be good. The date on your newest review has quietly become one of the things that decides whether you get the call.
For a couple of years the whole push was about getting reviews at all. Claim your page, gather some stars, get named when a customer asks ChatGPT for help. That advice is settled and it still holds. The part almost nobody acts on is that reviews age. An assistant deciding which dentist to suggest this morning is reading for signs of right now: is this place still open, still busy, still making people happy this month. A pile of glowing reviews from eighteen months ago answers a different question than the one the customer is asking.
Two reports point at the same date
I had been watching this play out in dental and home-service audits for months before either report landed. Two reports out this spring put numbers on it, one looking at what customers do and one at how the assistants choose.
On the customer side, Yext's Consumer Search Behaviors Report 2026 measured what a customer does after an AI assistant hands them a recommendation. The top thing that decides whether they act is the star rating, at 34%. Word of mouth comes next at 30%. Then how recent the reviews are, at 29%, ahead of how the reviews read and how many there are. Review signals fill five of the top six spots. So after the assistant says your name, the freshness of your reviews, meaning the date on the newest ones, is near the front of what moves the customer toward picking up the phone.
Birdeye's State of AI Search 2026 looked at the other end: how the assistants themselves choose. When they pick a local business to recommend, they weigh how recent your reviews are, whether your listing details are correct, and whether feedback keeps coming in steadily. To the assistant, a recent and ongoing stream of reviews signals a business that is currently active and well liked. If that stream dried up months ago, the same assistant treats the page as a business that may have slowed down.
A business with a five-star wall that stopped growing last winter looks, to both the machine and the person checking its work, like a business that used to be good. The one that quietly gathered four reviews last month looks alive (fewer in total, but visibly current).
The customer is going to check, and almost all of them do
Getting named by the assistant is only the first step, and the next one is where you can quietly lose the customer you already won. After it names you, the customer goes and checks before they act.
Yext found that more than 93% of AI users take at least one step to verify a recommendation before they go. 62% search Google. 58% go straight to the business website. 52% click the sources the assistant showed under its answer. That is nearly everyone, doing some version of the same thing: looking you up to make sure you're real and current.
So picture the customer who got your name from ChatGPT. She opens your Google listing to see for herself. She sees four and a half stars, which is good. Then she notices the newest review landed nine months ago. A small doubt creeps in. Are they still this good? Are they even still busy? She doesn't know your business the way you do. All she has is what the page shows her, and the page is telling her this place went quiet. That doubt is enough to send her to the business down the street whose newest review is from last Tuesday.
The 30-second test you can run this morning
Finding out where you stand takes about a minute. Go and look.
Pull up your own business the way a stranger would. Search your name on Google and open your Google Business Profile, which is the free listing that shows your reviews, hours, and map pin. Find your most recent review and read the date. That date is the whole test. If it's from the last few weeks, you're in good shape. If it's more than a month or two old, you just found the gap, and it's a thirty-second thing to start closing today.
While you are there, do the same on the main review site for your trade. A dental practice might check Healthgrades. A law firm would look at Avvo. A home-service contractor would open the directory its customers already trust. Read the date on the newest one. If those pages went quiet too, that is the same gap in another place a customer will look.
The fix is small, boring, and weekly
The cure is almost too plain to feel like one. Ask two or three happy customers for a review every week. Every week. That keeps a fresh review on top at all times, which is the whole point.
Make the ask easy. The best moment is right after good work, while the customer is still pleased and standing in front of you. Hand them a card with a short link, or text the link before they've left the parking lot (the ask lands best while the good mood is still fresh). The easier you make it, the more of them actually do it. Review platform guidance in 2026 puts the practical target at a handful a month, five to ten, which works out to two or three a week. That cadence keeps a steady flow of reviews coming in.
Two honest cautions. First, don't buy reviews or post fake ones. That breaks the rules of Google and most review sites and can get your listing penalized, and assistants are getting better at spotting it. Real reviews from real customers are the only version that works. Second, treat this as an ongoing habit. A burst of twenty reviews this week and then silence for a year leaves you right back where you started, with a newest-review date that keeps aging. The slow steady drip is what wins, because the thing you are protecting is the date.
Why this is worth a few minutes a week
Almost half of US adults, 47%, used an AI assistant to find a local business in the past month, by Yext's count. Enough of your future customers now start with ChatGPT or Gemini (the way they used to open a Google search) that the date on your newest review has started to affect whether your phone rings. That's the change worth a few minutes a week.
So this morning, look yourself up and read the date on your newest review. If it's older than a month or two, start the habit today: ask your next two happy customers for a review, make the link easy, and keep it going. Then check the date again in a few weeks. A business that always has a fresh review on top stays on the list the assistant recommends and passes the check the customer runs, both from the same small weekly habit.
Frequently asked questions
- Do old reviews stop ChatGPT from recommending my business?
- They make it less likely over time. Birdeye's 2026 report says assistants weigh how recent your reviews are when they pick a local business. A page that stopped collecting reviews months ago reads as a business that may have slowed down, so the assistant leans toward one that is still getting fresh feedback.
- How recent does my newest review need to be?
- Aim to always have one from the last few weeks. There is no exact cutoff, but if your most recent review is more than a month or two old, the gap is wide enough to notice. A handful of new reviews a month, roughly five to ten, keeps a fresh one on top at all times.
- Should I buy or fake reviews to keep them fresh?
- No. Buying or faking reviews breaks the rules of Google and most review sites and can get your listing penalized. The only safe way is to ask real customers right after good work. The fix only works if the reviews are real.