How AI search chooses local businesses, and what it means for your Google reviews

A growing number of people no longer scroll a list of results to find a plumber, a cafe or a hairdresser. They ask an AI assistant, or they read the short summary Google writes at the top of the page. If you run a local business, the obvious question is: how does it decide who to recommend, and is there anything you can do about it?

This guide sticks to what is actually known. A lot of what gets written about AI and local search is guesswork presented as fact, so where something is an observation and not a published rule, we say so.

What has changed

BrightLocal's 2026 Local Consumer Review Survey asked 1,002 US adults how they find local businesses. Use of ChatGPT and similar tools for local recommendations went from 6% in 2025 to 45%, which made AI the third most-used source. It is a US survey, so New Zealand numbers will differ, but the direction is hard to ignore.

What people do with those answers matters just as much:

• They still check reviews. 97% of consumers read reviews for local businesses, and 41% say they always do, up from 29% a year earlier.
• Standards have risen. 31% will only use a business rated 4.5 stars or higher (17% a year earlier), and 68% want four stars or more.
• Recent reviews count. 74% look for reviews from the last three months, and 32% look for reviews from the last two weeks.
• People read the summaries. 82% read AI-generated review summaries, and 23% would decide on the summary alone.
• Trust is mixed. 40% trust AI platforms for business recommendations and 32% do not. 42% trust AI recommendations as much as reviews.

So AI is becoming a new front door, and the reviews behind it are still what people check before they choose.

What Google actually says

Google does not publish a formula for who appears in AI answers, and it is clear about two things.

There is no special trick. Its documentation on AI Overviews and AI Mode says there are no additional requirements to appear in them. It points site owners to the same basics as ever: pages Google can crawl, good internal links, a good page experience, important content in text, and an up-to-date Business Profile. It also says you do not need special AI files, markup or structured data.

Local ranking rests on three things. Google's help page says local results are based mainly on relevance, distance and prominence. Prominence depends partly on how many reviews a business has and how many websites link to it. More reviews and positive ratings can help a business's local ranking, and Google recommends replying to reviews. It also states that there is no way to request or pay for a better local ranking.

Notice what that means. Nobody can buy a place in the answer, and none of it requires technical wizardry. The inputs are your Google profile, your reviews, and your presence on the web.

What outside researchers have seen

Beyond Google's own pages, agencies and researchers have studied AI answers. Treat these as informed observations, not rules, because the systems change often.

• AI Mode draws on Business Profile details. Practitioners report that it shows ratings, review counts, hours and services taken from Google Business Profiles, and that the businesses that show up often overlap with the usual map results.
• Other review sites can matter too. A Whitespark study looked at which review sites Bing Places draws on across 153 searches in nine US cities, and found Facebook, Yelp and Tripadvisor cited most often, depending on the category. How much ChatGPT relies on Bing is not settled, and the researchers themselves are cautious about it.
• Be wary of precise targets. You will see claims such as "you need 200 reviews to be recommended". We could not find any study behind numbers like that, so we would not plan around them.

What a small business can do

None of this needs a new strategy. It rewards the same habits that have always helped local businesses, and punishes shortcuts.

1. Keep your Google profile complete and current

Categories, services, opening hours, service areas, photos and a proper description. This is the information AI answers draw on, and it is the "relevance" part of Google's three factors. If you are not sure where to start, see how we set up and tune profiles.

2. Get a steady flow of new reviews, not occasional bursts

Recency is what people say they look for, so a handful of reviews every month beats a big batch once a year. The reliable way is to ask every customer, at the right moment, with a link that opens the review box in one tap. Our guides on asking customers and the best time to ask cover how. You can make your link and QR code with the free review link generator.

3. Ask everyone, and never pay for reviews

It is tempting to shortcut with rewards or to ask only your happiest customers. Both break Google's rules and risk reviews being removed, and AI summaries are built from the same reviews, so a distorted set of reviews gives a distorted summary. See review gating and incentives.

4. Reply to your reviews

Google recommends it, and it is also part of what people read. A short, specific reply that names the job and says thanks shows the next customer how you treat people. Our guides on replying to positive and negative reviews give templates.

5. Invite detail in reviews

This one is a sensible inference, not a proven rule. A review that says "they replaced our hot water cylinder the same day and left it tidy" carries more information than "great service", for the reader and for any summary built from it. You cannot script what customers write, and you must not coach them, but asking "if you can mention what we did for you, it helps other people decide" is honest and gentle.

6. Keep your details the same everywhere

Make sure your name, address, phone number and website match across your site, your Google profile and any directories. This is ordinary local SEO, and it removes confusion for any system pulling information about you.

7. Check other places people look

Since some AI tools draw on other sources, it is worth claiming your listing on Bing Places and checking your Facebook page is accurate. For most New Zealand local businesses, Google is the place to put your effort first.

What to ignore

• Anyone selling a guaranteed place in AI answers. Google says there is no way to buy a better local ranking, and no special file or markup gets you in.
• Fixed review-count targets. More recent, genuine reviews help. There is no magic number.
• Panic. The people choosing between you and a competitor are still reading reviews and still weighing a star rating. The basics have not moved.

How to tell if it is working

Ask new customers how they found you, and record it. Watch your Google profile's views and calls. Once you have a website, the Performance report in Google's Search Console includes traffic that comes through AI features. Over a few months, a rising number of recent reviews and a steady rating are the best signs you are on the right track.

Where Goodshout fits

Goodshout is built for the habit that matters most here: asking every customer, right after the job, with a one-tap link to your Google review box, and a private way to tell you if something went wrong. It does not promise a ranking, because nobody can. It makes the reliable part, a steady flow of recent, honest reviews, something that happens without you having to remember.

Sources: BrightLocal's Local Consumer Review Survey 2026, Google Search Central's AI features and your website, Google's help page on local ranking, and Whitespark's research on review sites and Bing Places. Checked October 2026. Survey figures are from a US sample.