In its 2026 Local Consumer Review Survey, BrightLocal found that 47% of customers will not use a business with fewer than 20 reviews. Only 9% would use one with five or fewer.
Eleven reviews is under the line for about half the people who look you up, before any of them read a word. Most owners know reviews matter and have no idea what number is enough. There is an answer, and it has three parts.
The three thresholds
Volume comes first, and 20 is the number to clear.
Your star average is second, and the bar sits at 4.5. In that same BrightLocal survey, 31% of customers said they will only use a business rated 4.5 stars or higher, and 68% said 4.0 or higher.
Recency is third, and the window is 90 days. BrightLocal’s 2026 numbers put 74% of customers in the group that only cares about reviews from the last three months.
Miss one and the other two stop carrying you. Forty reviews at 4.8 stars, all from 2023, is a page that says you used to be good.
Why the star rating bar moved 13 points in a year
BrightLocal’s 2026 survey has 68% of customers refusing to use a business rated under 4.0 stars, up from 55% a year earlier, and 31% holding out for 4.5 or better, up from 17%.
The survey does not say why. It does show, in the 2025 edition, that 42% of customers now trust online reviews as much as a personal recommendation, down from 79% in the 2020 version of the same survey. Trust in the medium fell by more than half while the standard people hold it to went up.
Settle one thing first. Nearly every figure here comes from one annual survey that BrightLocal, a local SEO software company, runs on about 1,002 US adults. BrightLocal sells review management software, so research concluding that reviews decide purchases is good for its business and for ours. Read the direction as real and the decimals as soft. A move from 55% to 68% counts. A two-point difference is noise. (The Pew and Clutch figures further down come from outside BrightLocal.)
Recency, and why a one-time review push fails
Here is the version most owners try. You email your customer list in March, collect 25 reviews in ten days, and feel caught up. By July those reviews sit outside the window three quarters of your customers care about. BrightLocal also found 32% of customers look specifically for reviews from the last two weeks, up from 20% in 2025.
The arithmetic is ours, not BrightLocal’s: holding 20 reviews inside a rolling 90-day window takes about seven new ones a month. For a shop doing 30 or 40 jobs a month, that is one customer in five saying yes. It is a weekly habit.
Where they land matters less. BrightLocal found 74% of customers in the 2025 survey used two or more sources. Google first, then one other platform that carries weight in your trade.
One place reviews do almost nothing is your own website. Clutch asked 263 US consumers in June 2025, a small sample, what makes a website feel secure. Testimonials or reviews came in at 4%, behind an SSL lock icon at 33% and clean professional design at 23%. Reviews decide the purchase on the platforms customers look for them on. Pasted onto your own home page, they barely register. Worth knowing before you pay somebody to build a testimonial carousel.
The response expectation, and the template trap
BrightLocal’s 2026 survey is blunt here. 89% of customers expect owners to respond to reviews, 80% are more likely to use a business that responds to every one, and 42% are unlikely to use a business that ignores its reviews entirely.
Then the finding that wrecks the standard advice: templated or generic responses make 50% of customers unlikely to choose a business. Half. Nine reviews in a row answered with “Thank you for your feedback!” is visible from across the room.
What works takes about a minute. Name the thing the customer mentioned, add one detail only somebody who was there would know, sign it with a real first name:
Thanks, Dana. Glad the new gutters held through that stretch of rain in June. Call the shop if the downspout on the north side ever backs up again. Greg
BrightLocal found 19% of customers expect a response the same day they post, up from 6% the year before, and 81% expect one within a week. One business day is a defensible standard for a small team. Answer the bad ones too, without arguing in public.
How to ask without it feeling like begging
The awkwardness comes from treating the ask as a favor. It is closer to paperwork.
Ask at the moment the customer says something good out loud: the walk-through, the handoff, the follow-up call. Not on the invoice, and not two weeks later. Then send the link by text.
In person:
Would you be willing to leave us a Google review? I’ll text you the link right now. Takes about a minute, and it’s mostly how people around here find us.
By text within the hour:
Hi Dana, it’s Greg with [business]. Thanks again for having us out for the gutter work. If you have a minute, a short review helps your neighbors find us: [link]. If you’d rather skip it, no hard feelings.
One follow-up three days later, one line long, and then you stop. Two messages total. You asked in person first, so the text is expected, and you gave them an exit in writing. That is what keeps it from feeling like begging.
Now what we will not do for you. No review gating, meaning surveying customers first and sending the link only to the ones who answered happily. That violates Google’s policies and it is lying about your own business with extra steps. No buying, swapping or writing reviews, and no incentives, which break the platforms’ rules.
Ask everybody, including the customer you suspect is a little annoyed. The instinct to screen is understandable, and it is still against the rules.
What AI review summaries change
A separate BrightLocal report from March 2026 found 82% of customers read AI-generated review summaries. 23% will decide on the summary alone, 59% use it as a starting point and then check the rating and the full reviews, and 18% skip summaries entirely.
Before you rearrange your week around that, check how many people are in this funnel at all, using a source with no stake in review software. Pew Research Center’s June 2026 report on Americans and AI surveyed 5,119 US adults in February 2026 through its American Trends Panel, recruited by address-based national random sampling. That is a better-sourced number than any vendor survey in this post. Pew found 49% of US adults ever use an AI chatbot, up from 33% in 2024, and that looking up information is the most common use at 42%. So about half your customers are anywhere near an AI summary, and half are not.
You cannot write that summary. It is assembled from what your customers wrote. Twenty reviews that all say “Great service!” give the summarizer nothing. Twenty that name the job, the town and the person who showed up give it something to say about you.
Two more numbers from the 455 people in BrightLocal’s 1,002-person panel who had used an AI tool to find a local business: 88% check the reviews an AI cites, and 71% trust AI to summarize real reviews accurately, against 50% of customers overall. AI is a filter in front of your reviews, not a substitute for them.
The monthly rhythm
None of this takes long. It has to happen on schedule.
Daily, five minutes. Check for new reviews and answer them, same day where you can. That covers the 19% who expect same-day and keeps you inside the week 81% expect.
Weekly, fifteen minutes. Go through the jobs you closed that week and send the ask to each one. Seven a month is the number that holds 20 reviews inside the 90-day window.
Monthly, twenty minutes. Count the reviews posted in the last 90 days and compare that to 20. Check your star average against 4.5 and 4.0. Search your own business name and read the AI summary that comes up.
Two hours a month. What kills it is never the two hours. It is that the routine lives in somebody’s head instead of on a calendar.
That is a system rather than a campaign, and a system quietly stops running the month you get busy. Which is most months. When we handle this for a client, the shape is exactly what is written above: somebody other than you watches the count and the window and answers reviews inside a day.
Reviews you earned, from customers you served, answered in your own words. That is the only version that holds up, and the only one the platforms permit.
Start with the count. Open your Google Business Profile, count the reviews posted in the last 90 days, and write down your star average. Two numbers, five minutes. If the count is under 20, the answer is the weekly habit described above and nothing cleverer. If the average is under 4.5, read your last ten reviews and the responses under them before you change anything else. Then check both numbers again on the first of next month.