Ninety-seven percent of consumers read reviews for local businesses, and 41% now say they always do when browsing, up from 29% a year earlier. Both figures come from BrightLocal’s Local Consumer Review Survey 2026, a panel of 1,002 US adults published in February 2026.
The first number has been roughly stable for years and is quoted everywhere. The second one moved twelve points in twelve months, and almost nobody is quoting it. That gap is typical of this research area: the headline statistics are stale and comfortable, and the movement is in the figures underneath them.
What follows is the set of consumer trust statistics actually worth carrying into 2026, each with its source, plus the four that should change what you do this week rather than what you say in a deck.
Almost Everyone Reads Reviews, and the Habit Hardened
The reading behaviour is now close to universal, so the interesting variable is intensity rather than reach.
Alongside the 97% who read reviews and the 41% who always do, consumers consult around six review sites on average before deciding. That last figure is the one that undermines most single-platform reputation strategies. A business with a strong Google profile and nothing elsewhere is answering one of six questions a prospective customer is asking.
The platform mix also shifted sharply. Google review usage fell from 83% to 71% in a single year. Apple Maps rose from 14% to 27%. Local news sites dropped from 48% to 29%. None of those are small movements, and the Google decline is the one most businesses would assume could not happen.
Two outcome figures anchor why any of this matters. Eighty-five percent say they are more likely to use a business after reading positive reviews, and 77% say they are less likely after negative ones. Ninety-three percent have made a purchase after reading reviews.
One figure here deserves separating out because it is the most quoted and the most slippery. Forty-nine percent say they trust reviews from strangers online as much as personal recommendations. That is a striking number and it is commonly rendered as “half of people trust reviews as much as friends,” which is close enough to be fair. What it does not mean is that reviews substitute for word of mouth in aggregate, because the half who do not trust strangers equally are still making decisions, and they are the half who will ask someone they know. Both channels are live, and a business that is strong in reviews and absent from referral conversation is reaching one half of the market well.
The Star Rating Bar Jumped in One Year
This is the most actionable movement in the data and the least discussed.
Ninety-two percent of consumers say they care about star ratings. The threshold they apply has tightened fast. Sixty-eight percent will only use businesses rated 4.0 or above, up from 55% the previous year. Thirty-one percent now require 4.5 or better, nearly double the 17% recorded a year earlier. Ten percent will only use five-star businesses.
Read the 4.5 figure again, because the rate of change is the story. A business sitting at 4.3 was comfortably above the bar for most consumers last year and is now excluded by roughly a third of them. Nothing about the business changed.
Volume thresholds matter too, and they are more forgiving than people fear. Forty-seven percent will not use a business with fewer than 20 reviews, and only 9% will use one with five or fewer. Twenty is therefore the real floor, and it is reachable for almost any business within a quarter of asking properly.
Recency Now Beats Volume
The recency figures are the ones that change how you operate rather than what you aim for.
Seventy-four percent look for reviews from the last three months. Thirty-two percent want reviews from the last two weeks, up from 20% a year earlier. Eighteen percent say they are only swayed by reviews from the last week.
Put that next to the volume data and a clear operating rule falls out. A profile with 300 reviews and nothing in the last two months is weaker than one with 40 reviews and four from last week, for a third of the market. Review generation is not a project you complete. It is a cadence you maintain, and the required cadence is now roughly weekly for a business of any size.
What customers do after reading a positive review is worth knowing too, because it sets what the review has to accomplish. Sixty-six percent do further research. Fifty-four percent visit the business website, up from 32% in 2019. Thirty-seven percent read more reviews, 31% visit the location, and 34% say they are ready to buy or book.
That first number is the one to design around. Two thirds of people who liked what they read go looking for more, which means the review is a referral to your own site rather than a close. If your reviews are strong and your site is weak, you are generating traffic that converts worse than it should, and the review profile will get the credit for a problem it did not cause.
The factors consumers say they weigh inside an individual review reinforce this. Fifty-six percent cite a review being backed by other reviews with similar sentiment, 44% cite it being posted within the last month, and 42% cite a high star rating. Recency outranks the rating itself.
AI Became the Third Channel in Twelve Months
This is the largest single-year shift in the dataset and the reason the rest of your reputation work needs rethinking.
Forty-five percent of consumers used AI tools for local business recommendations in 2026, up from 6% the year before. That places AI third among discovery channels, behind Google and Facebook and ahead of Yelp and TripAdvisor. ChatGPT accounted for 31% of recommendation use, with Google’s AI Mode at 23%.
Age is the main divider. Adults aged 30 to 44 lead at 64% usage. People over 60 trail at 24%. If your customers skew to the middle bracket, nearly two in three are already asking an assistant.
Trust in those recommendations is higher than sceptics expect. Forty percent of all consumers say they trust AI platforms for business recommendations, against 32% who do not. Forty-two percent trust AI recommendations as much as traditional reviews. Among the subset who actually use AI tools for this, a sample of 455 respondents, trust rises to 63%, with only 10% expressing distrust.
Review summaries are now a layer in their own right. Eighty-two percent read AI-generated review summaries, 23% would rely on the summary alone, and 18% skip them. Half of all consumers trust AI platforms to accurately summarise reviews from real people, rising to 71% among AI users.
There is a behavioural check in the data that should temper the panic slightly. Eighty-eight percent of AI users say they fact-check AI outputs, with 51% checking whether a review is legitimate and 37% checking the source, while 12% do not check at all. So the assistant’s answer is rarely the final word for an engaged buyer.
The risk is not that an AI answer goes unverified. It is that you never make it into the answer that gets verified. Being absent at stage two of the loop means the verification stage happens on a competitor’s profile instead of yours, and no amount of review quality rescues a business that was not mentioned.
The operational consequence is uncomfortable. For a meaningful slice of your market, the first description of your business is written by a model summarising your reviews, and you have no editorial control over it. What you have control over is the input.
Check Whether You Pass the Response Test
Four figures here, and together they make review responses a commercial function rather than a courtesy.
Eighty percent are more likely to use a business that responds to all its reviews. Forty-two percent are unlikely to use one that never responds. Eighty-nine percent expect business owners to respond at all.
Speed expectations have compressed. Nineteen percent now expect a same-day response, up from 6% a year earlier. Thirty-two percent expect one by the next day, up from 18%. Eighty-one percent expect a response within a week.
And then the figure that should stop anyone about to automate this: 50% say they are unlikely to choose a business that uses generic or templated responses. Half the market treats a visibly automated reply as worse than silence, which inverts the logic of most review management tooling.
There is also an asymmetry worth knowing. Forty-five percent are less likely to use a business that only responds to positive reviews, and 47% if it only responds to negative ones. Selective responding reads badly in both directions.
The Verification Loop
Here is the framework that pulls these consumer trust statistics into something you can act on. Call it the Verification Loop, and it describes the path a 2026 customer actually takes.
The loop has four stages. A customer asks an assistant for a recommendation. The assistant answers from summarised review and web content. The customer then verifies the answer on a native review platform, because 97% of AI users say they sometimes double-check recommendations against real reviews and 42% always check native platforms. Then they decide.
What matters is that the loop runs on the same underlying material twice. Your reviews generate the AI answer at stage two and then get read directly at stage three. A thin or stale review profile fails at both points, from one cause.
This also explains why the old split between search optimisation and reputation management has stopped being useful. They are now the same input. The text of your recent reviews is simultaneously your ranking signal, your AI summary source and the thing a verifying customer reads with their own eyes.
The practical test is to run the loop yourself. Ask ChatGPT for the best provider in your category and city, see whether you appear, then read what it says about you and check which reviews it drew that from. Most businesses doing this for the first time find the description is accurate and three years out of date, which is a fixable problem and an invisible one until you look.
Cite These Carefully
Three rules, because this is an area where sloppy citation gets caught.
Name the study and year in the sentence. “BrightLocal’s Local Consumer Review Survey 2026” is five words and removes the main objection. The survey polled 1,002 US adults via a SurveyMonkey panel and published in February 2026, so it is US-specific and panel-based, and both caveats are fair to state.
Keep the AI figures separate from the general ones. Several of the most striking numbers come from the 455-respondent subset who actually use AI tools, not from the full sample. The 63% trust figure is an AI-user figure. The 40% is the all-consumer figure. Conflating them inflates the claim and the difference is checkable.
And treat the fake review numbers as a separate subject. Ninety-seven percent think businesses caught with fake reviews should be punished, 57% say they should be banned from review platforms, and Google reported blocking over 240 million fake or policy-breaking reviews in 2024. Those belong in a conversation about enforcement, not in a slide about trust.
The Four That Should Change Your Week
Most of the numbers above are context. Four of them are instructions.
The 4.5-star threshold doubled to 31% in a year. Find your current rating, and if it sits between 4.0 and 4.5, closing that gap is the highest-return reputation work available to you, because a third of the market is applying a bar you are just under.
Thirty-two percent want reviews from the last two weeks. Set up a weekly ask rather than a quarterly campaign. The mechanism matters less than the rhythm.
Fifty percent are put off by templated responses. If you are running automated replies, switch to writing them, and if the volume makes that impossible, write them for a subset and leave the rest unanswered rather than templated.
And 45% now use AI for local recommendations, up from 6%. Run the Verification Loop on yourself this week, note what the assistant says about you, and treat the gap between that description and reality as your actual reputation problem.
Expect the AI figure to keep climbing, which means the review text you generate this quarter is writing the description a stranger will read next year.