How many people read reviews, and where
Almost everyone reads reviews before choosing a local business: 97% of US consumers in BrightLocal's 2026 survey. What changed in a year is how often. The share who always read reviews rose from 29% to 41%.
Google is still the main place people read them, but its share fell from 83% to 71% as AI assistants took off. 45% of consumers now use ChatGPT or similar tools to find reviews, and those tools mostly summarise the same Google, Yelp and Facebook reviews. A thin or stale profile shows up there too.
Show the data
| 2025 | 2026 | |
|---|---|---|
| Always read reviews before choosing a local business | 29% | 41% |
| Use Google to read reviews | 83% | 71% |
| Use ChatGPT or other AI tools to find reviews | 6% | 45% |
| Use Apple Maps to read reviews | 14% | 27% |
Source: BrightLocal Local Consumer Review Survey 2026 (n=1,002 US adults)
The star rating customers expect
The bar went up. Two thirds of consumers (68%) won't consider a business under 4.0 stars, up from 55%. Nearly a third (31%) want 4.5 stars or more, almost double the 17% of 2025.
Volume and freshness matter as much as the average. Almost half of consumers skip a business with fewer than 20 reviews, and three in four only trust reviews from the last three months. A 4.8 built on twelve reviews from 2023 doesn't do the job.
Show the data
| 2025 | 2026 | |
|---|---|---|
| Under 4.0 stars | 55% | 68% |
| Under 4.5 stars | 17% | 31% |
Source: BrightLocal Local Consumer Review Survey 2026 (n=1,002 US adults)
| What they check | Share of consumers |
|---|---|
| Care about the star rating at all | 92% |
| Won't use a business with fewer than 20 reviews | 47% |
| Look for reviews written in the last 3 months | 74% |
| Look for reviews from the last 2 weeks (20% in 2025) | 32% |
| Only consider reviews from the past week | 18% |
Source: BrightLocal Local Consumer Review Survey 2026 (n=1,002 US adults)
What a better rating is worth
Small rating changes move real numbers. The effect is strongest around the thresholds above: moving from a 3.x to a 4.x rating, and getting past the first handful of reviews.
| Finding | Effect | Study |
|---|---|---|
| Rating goes up by 0.1 star | +25% conversion | Uberall, 64,000 locations (2019) |
| Rating goes from 3.5 to 3.7 stars | +120% conversion | Uberall, 64,000 locations (2019) |
| Product shows 5 reviews instead of none | 270% more likely to be bought | Spiegel Research Center (2017) |
| Where purchase likelihood peaks | 4.0–4.7 stars, not 5.0 | Spiegel Research Center (2017) |
| One extra Yelp star for an independent restaurant | +5–9% revenue | Michael Luca, Harvard Business School |
Asking for reviews works, and the channel matters
Most customers leave a review when they're asked. 78% of consumers were asked in the past year and 83% of those wrote one. The share who always write a review when asked rose from 16% to 28%.
How you ask changes the result. In GatherUp's customer data, email requests produced about 15 reviews per 100 sent, SMS about 20, and businesses using both got 26. Whatever the channel, ask every customer: asking only the ones you expect to be happy breaks Google's rules.
Show the data
| Reviews per 100 requests | |
|---|---|
| Email only | 15 |
| SMS only | 20 |
| Email and SMS | 26 |
How fast customers expect a reply
Replying is no longer optional, and the clock got shorter. 19% of consumers expect a reply to their review the same day, up from 6%, and 32% expect one by the next day. 81% expect one within a week.
Templates are noticed: 80% are likely to use a business that responds to all its reviews, but 50% are put off by generic, copy-paste replies.
Show the data
| 2025 | 2026 | |
|---|---|---|
| The same day | 6% | 19% |
| By the next day | 18% | 32% |
Source: BrightLocal Local Consumer Review Survey 2026 (n=1,002 US adults)
Fake reviews and Google's crackdown
Google blocked or removed 292 million policy-violating reviews in 2025, up from 240 million in 2024 and more than five times the 55 million of 2020. It also removed 13 million fake Business Profiles, the same as the year before, and now uses Gemini to catch review manipulation automatically.
The same systems look for review gating, incentives and staff quotas. A profile caught can have new reviews frozen and existing ones removed, which undoes years of collecting them.
Show the data
| Reviews removed | |
|---|---|
| 2020 | 55M |
| 2021 | 95M |
| 2022 | 115M |
| 2023 | 170M |
| 2024 | 240M |
| 2025 | 292M |
Source: Google's yearly Maps safety reports, via Search Engine Roundtable and IANS
What to do with these numbers
- 1
Ask every customer, once, soon
83% of asked customers leave a review, so the biggest lever is asking at all. Send one request shortly after the visit, to everyone.
- 2
Keep reviews coming every month
74% only trust reviews from the last three months. A steady trickle beats a one-off push.
- 3
Reply within a day
Half of consumers expect a reply by the next day. Write a real reply, not a template.
- 4
Watch the 4.0 and 4.5 lines
Those are the thresholds where customers drop you from their shortlist. Know how many 5-star reviews it takes to cross the next one.