
Short answer: In a sample of 9,435 Google AI Mode answers to home services questions, the businesses it named had a median rating of 4.9 stars, 83% were rated 4.8 or higher and only 1.0% were under 4.0. Review counts were far more forgiving: 8% of the named businesses had fewer than 20 reviews. The takeaway for a local business is that a weak star rating is a bigger obstacle than a short review history. The catch is that this is one company's measurement of who got named, not a rule Google has published.
If you run a plumbing, HVAC, electrical or roofing company in Fort Wayne, Auburn or anywhere in Allen and DeKalb counties, the question is practical: when a homeowner asks an AI tool who to call, what do the businesses it names have in common, and what can you change? This post walks through the numbers, what they do and do not prove, and what we would fix first.
A note on sources. The data comes from Cheers, a company that sells AI visibility tools, and we read its article in full. We also read Google's own guidance on generative AI search, Google's review policy, and a critique of the AI search sales pitch from Blacksheep Creative. We did not run this study ourselves, and we sell local SEO services, so weigh our advice with that in mind.
Key Takeaways
- Cheers collected 9,435 Google AI Mode answers to 32 home services questions in 100 US and Canadian metros between August 31 and September 20, 2026. About 90% of the answers showed rated business cards.
- Across 34,017 rated business appearances, the median rating was 4.9 stars, 83% were 4.8 or higher and 1.0% were under 4.0. The median review count was 226.
- Only 3.7% of answers named any business under 4.0 stars, but 23.5% named at least one with fewer than 20 reviews. The rating floor looked firmer than the review floor.
- When an answer named three or more businesses, the first one had a median of 342 reviews, against 237 for the second and 195 for the third. Cheers calls that a correlation only.
- Google says generative AI search is rooted in its core ranking and quality systems, that no special markup is needed, and that Business Profiles can help visibility. Ask every customer for reviews the same honest way, because Google's policy bans incentives and selective asking.
What Did the Study Actually Measure?
Every week, Cheers asks ChatGPT, Gemini, Perplexity and Google AI Mode the same 32 home services questions in 100 metro areas across the United States and Canada, and records every business each engine names, with its star rating, review count and every source it cites. The article we are using covers three weeks, from August 31 to September 20, 2026, and 34,742 usable answers across the four engines. Of those, 9,435 came from Google AI Mode, and those are the only numbers we use here.
The questions covered plumbing, HVAC, electrical and roofing, and they came in several flavors: best company for a job, emergencies, same-day requests, questions about reviews and fair pricing, and requests for deals and quotes. The sample spans 60 US and 40 Canadian metros. The article we read does not list the cities, so we cannot tell you whether Fort Wayne was one of them.
Cheers is careful about what it claims, and it is worth repeating its warning before any numbers appear. In its words:
Every number here describes the businesses Google AI Mode named, not a rule Google applies. Google publishes no local recommendation formula, so read each finding as an association in our sample.
That sentence is the right frame for everything below. A study can tell you what the named businesses look like. It cannot tell you what Google would do with your business if you changed one thing.
Who Did Google AI Mode Put on Its Cards?
A Google AI Mode answer to a home services question was, nine times in ten, a short list of business cards. Cheers found rated cards in 90% of the answers it collected, and the share barely moved by question type. A typical answer carried about four cards, and AI Mode showed a star rating and review count for 90% of the businesses it named.
| Type of question | Answers that showed rated business cards |
|---|---|
| Best company for this job | 89.7% |
| Emergencies | 91.6% |
| Same-day requests | 90.0% |
| Reviews and fair pricing | 90.9% |
| Deals and quotes | 93.3% |
The practical meaning is that for this kind of search, your star rating and review count are not a footnote. They sit on the card, next to your name, in front of the customer. The businesses on those cards looked like this:
median star rating of businesses AI Mode named with a rating (Cheers, 34,017 rated appearances in 8,523 answers)
of those appearances were rated 4.8 stars or higher (Cheers)
were rated under 4.0 stars (Cheers)
median Google review count; 8% had fewer than 20 and 21% had 1,000 or more (Cheers)

Think of the illustration above as the sorting step. Plenty of local businesses could answer a homeowner, and the answer shows about four of them. Whatever Google does to pick, the four that came out the bottom were overwhelmingly rated in the high 4s. Review counts, as the numbers show, were spread much wider: the median was 226, but one named business in five had 1,000 or more, and one in twelve had fewer than 20.
A word on the rounding. Google AI Mode shows review counts above 999 in a rounded form, such as 2.0K, and its place cards hid business names in about 5% of answers, according to Cheers. Both are small details, and both mean the figures are approximations of what a customer saw.
Is the Star Rating a Firmer Line Than the Review Count?
This is the finding we think matters most to a small business, because it changes where you put your effort. Cheers looked at the 8,523 answers that showed rated cards and asked how often an answer named at least one business below a given line:
- 3.7% of answers named at least one business rated under 4.0 stars.
- 14.6% named at least one rated under 4.5 stars.
- 23.5% named at least one business with fewer than 20 reviews.
Read those three lines together. A business with a low rating almost never showed up. A business with very few reviews showed up in about one answer in four. Cheers puts it this way: a thin review count did not keep a business off the list nearly as often as a weak rating did.

There are two sides to that, and Cheers spells both out. For a newer or smaller company, a short but strong history was still being named. For an established company that has slid into the low 4s, AI Mode rarely named businesses there. Cheers is honest that it cannot tell why: “We can't tell whether Google screens them out or whether those businesses simply lose on other grounds, since we only see who was named.”
The review floor also moved by trade and by country. In electrical answers, 40.8% named at least one business with fewer than 20 reviews, against 11.6% of HVAC answers, where the median named business showed 449 reviews. Canadian businesses showed fewer reviews than US ones, a median of 134 against 332. If your trade is crowded with companies that have thousands of reviews, a thin history may hurt you more than the headline number suggests. If your trade is quieter, it may hurt less.
What a rating climb actually takes
Here is the arithmetic, which is ours and not Google's. This is a hypothetical, not a client account. Say a company has 100 reviews averaging 4.2 stars. Each new five-star review adds 5 points to a running total of 420. To reach a 4.5 average, you solve for the number of new five-star reviews, n: (420 + 5n) divided by (100 + n) equals 4.5. That works out to n = 60. You need about 60 new five-star reviews in a row, and any review below five makes it longer.
A smaller company with 50 reviews at 4.4 faces a steeper path to 4.8: it needs 100 straight five-star reviews. The lesson is not to give up. It is that the cheapest rating point is the one you do not lose. Whatever pattern keeps producing three-star reviews, such as missed callbacks, surprise pricing or no-shows, is the thing to fix before you ask for more.
Does the First Business Named Have the Most Reviews?
Order carried a signal the rating did not. In answers that named three or more rated businesses (7,873 of them), the first business named had a median of 342 reviews. The second had 237 and the third had 195. The most-reviewed business came first 36% of the time, where chance alone would put it first 25% of the time. Average ratings barely moved from first place to third.
median reviews for the business named first (Cheers, 7,873 answers)
median reviews for the business named second (Cheers)
median reviews for the business named third (Cheers)
Cheers flags this as a correlation across answers, and it is right to. Review count also tracks how big and how old a business is, and bigger, older businesses may be named first for reasons that have nothing to do with reviews. Its read for an operator is that once the rating clears the high 4s, review depth is the clearest difference between the first name and the third.
We would take that as a reason to keep collecting reviews steadily, not as a promise that more reviews will move you up the list. A business that goes from 120 to 240 reviews over a year by asking every customer is in a stronger position for many reasons. AI Mode might be one of them.
What Else Did AI Mode Lean On Besides Reviews?
Reviews were only part of what the study recorded. A few other findings change what a contractor should check this month.
- Google's own properties and your website did most of the citing. AI Mode cited Google's own properties in 95.5% of its answers and businesses' own websites in 92.5%. Only 38.3% of its answers cited any directory, review platform or “best of” list, against 84.2% for ChatGPT. For AI Mode specifically, your profile and your site matter more than they do for ChatGPT, where we covered a different pattern in our post on ChatGPT local results and Yelp.
- Most links to a contractor's site went to the homepage. When AI Mode linked to a contractor's website, 66.9% of those links went to the homepage and 13.5% to a location or city page. Service pages drew just 5.0% of the business-website citations. Cheers thinks the homepage links come from the website button on the Business Profile, and calls that an inference.
- Emergency answers talk about hours. For emergency questions, 92.8% of AI Mode answers mentioned 24/7 or after-hours availability, against 48.0% for “best company for this job” questions. Cheers notes that this counts any mention, including generic advice, so it shows what the answer talks about, not why a business was picked.
- Booking links showed up in almost a third of answers. In 31.2% of AI Mode answers, a link went straight to a booking or scheduling page on field service software. The share was 39.4% for plumbing, 38.9% for HVAC and 33.3% for electrical, and only 1.8% for roofing. We covered how those links work in our post on Google AI Mode booking links for home services.

The mockup above is a simplified picture of what a customer sees: a short answer, then a stack of cards with stars and a button. Everything on a card is something you can influence. The stars come from your customers. The button and the link come from your Business Profile. The hours come from the same place. And the page behind the link is yours. Our earlier post on the Book Online button Google may add on its own is a reminder to check those links yourself.
Is This Just Local SEO With a New Name?
Mostly, yes, and Google says so itself. Its guide to generative AI features in Search says the best practices for SEO remain relevant because the AI features are “rooted in our core Search ranking and quality systems.” On the new acronyms, the guide says that from Google Search's perspective, optimizing for generative AI search is optimizing for the search experience, and thus still SEO. It points owners to its guidance on evaluating third-party SEO advice if they are weighing “AEO” or “GEO” services.
The same guide says you do not need to create new machine readable files, AI text files, markup or Markdown to appear in Google Search, and that structured data is not required for generative AI search, with no special schema.org markup to add. On local, it says tools like Google Business Profiles can help products and services be visible in both AI responses and other Google Search results.
Clint Sanchez of Blacksheep Creative made the same point on November 9, 2026, with some heat. He wrote that the pitch he was seeing, a scary line about AI changing search followed by an offer of an AI audit, sells “the same website basics good agencies have been building for over a decade.” He accepts that the Cheers study behind such pitches is real. His objection is to the conclusion drawn from it.
We agree with most of that, and we would add one thing. The study does not say AI Mode changed what works. It says AI Mode is showing, in public, what a good local presence already looked like: a strong rating, a real review history, an accurate profile and a website that says what you do. That is a useful thing to measure, because it tells an owner where the work pays off. It is not a reason to buy a special AI package. If someone sells you one, ask what it changes that a solid Business Profile, a clear website and honest reviews do not.
If you want the wider picture, our post on how reviews affect SEO and AI visibility goes deeper on the review side, and our post on the map pack in AI Overviews covers how local searches still get a map.
How Do You Ask for Reviews the Way Google Allows?
If a 4.8 rating on a real review history is what the named businesses share, the temptation is to shortcut it. Google's policy closes that door. We read the Maps user generated content policy page on prohibited and restricted content, and here is what it says merchants may not do:
- Offer incentives, “such as payment, discounts, free goods and/or services,” in exchange for posting a review, or for revising or removing a negative one.
- “Discourage or prohibit negative reviews, or selectively solicit positive reviews from customers.”
- Require or pressure customers to leave ratings or write reviews while on the premises, or request that specific content be included.
- Request that staff solicit a certain number of reviews, or reviews that include specific content, including content that identifies a staff member.
What the policy does allow is the plain version: merchants may “solicit or encourage the posting of content that does represent a genuine experience, without offering incentives to do so or attempting to influence the rating or the contents of the review.” That is the whole recipe. Ask every customer, the same way, and let the rating be what it is.
The policy also bans content that shows “unusual volumes or patterns of review contributions that are indicative of efforts to manipulate a place's rating.” A sudden burst of reviews from a rushed campaign is the kind of pattern that line is about. Steady asking after each job looks different from a spike. And if a review does get flagged or removed, our post on Google review appeals running late explains what to expect.
What Should a Fort Wayne Service Business Do This Month?
We cannot tell you where Fort Wayne sits in the Cheers sample, and we would not guess. What we can say is that the steps below are cheap, they follow Google's own guidance, and they help whether or not AI Mode is the tool a customer used. This list is ours, built from the study and Google's pages. It is not Google's list.
- Look at your own card. Search your trade and city in Google, including AI Mode, and see what your rating and review count look like next to the businesses it names. Write both numbers down. That is your baseline.
- Read your last 20 reviews. Group the low ones by cause: late arrival, price surprise, no callback, messy work. Fix the operational problem first, because more reviews will not lift a rating that the service keeps earning.
- Ask every customer the same way. Pick one neutral request, such as a text with your Google review link after each job, and have every technician or front-desk person use it. No incentives, no filtering, no quotas for staff.
- Check the links and hours on your profile. The booking link should open a working page for your business, and the hours should match what happens when the phone rings after 6 p.m. Do not claim 24/7 service you cannot answer.
- Point your website button somewhere useful. If most AI Mode links land on the homepage, make sure your homepage says what you do, where you work and how to reach you. If you serve several towns, a page for each place you really serve gives people a better landing spot.
- Skip the special AI files and markup. Google says they are not needed. Spend that time on the items above.
If you run a dental office, a restaurant or a lawn care company, the numbers above are not yours. Cheers measured home services questions, and we do not have a comparable AI Mode sample for your trade. The steps still make sense. Your rating and review history are the first thing a customer sees, and reviews in healthcare carry extra privacy rules that we cover in our post on healthcare reviews and HIPAA.
What Does This Study Not Tell Us?
- Cause. Cheers sees who was named, not why. High ratings could be a screen, or they could simply travel with other things that matter, like a complete profile and a good website.
- Counting accuracy. Business names were extracted automatically. A hand check of 40 answers found the extraction caught about three in four of the businesses a reader would count, so Cheers calls its counts lower bounds.
- Repeat variation. Each question was asked once per city per week, so week-to-week changes include normal answer-to-answer variation. Cheers also says the businesses named change from week to week.
- Your trade and your town. The questions were about home services, and the trade figures cover plumbing, HVAC, electrical and roofing. The metros were a mix of US and Canadian cities.
- Where the data comes from. Cheers sells AI visibility services, so its numbers are one commercial company's measurements. We did not rerun them. We think they are useful, and we would not bet a marketing budget on any single percentage.
If you have checked your own business against AI Mode and found something that contradicts the pattern here, we would like to hear about it. One counterexample teaches more than another average.
Want a Second Look at Your Google Profile and Reviews?
Button Block helps Northeast Indiana service businesses tidy up their Google Business Profile, set up an honest review routine and make sure their website says what they do. We can walk through your profile, your last 20 reviews and the links customers click. We sell these services, so weigh our advice with that in mind.
Frequently Asked Questions
- What star rating do I need to show up in Google AI Mode?
- Nobody outside Google can give you a number, and Google publishes no local recommendation formula. In the Cheers sample of 9,435 AI Mode answers for home services, the median rating of the businesses named was 4.9 stars, 83% were rated 4.8 or higher and only 1.0% were under 4.0. That describes who got named. It is not a cutoff Google applies.
- Can a business with few reviews still get named?
- In that sample, yes. About 23.5% of the answers that showed rated cards named at least one business with fewer than 20 reviews, and 8% of all named businesses had fewer than 20. The pattern varied by trade: electrical answers named a thin-review business 40.8% of the time, against 11.6% for HVAC. A short, strong history was not disqualifying. A weak rating was a bigger problem.
- Does this study cover lawn care, dental offices or restaurants?
- No. The 32 questions were about home services, and the trade figures Cheers publishes come from plumbing, HVAC, electrical and roofing. We would not assume the same numbers hold for lawn care, a dental practice or a restaurant. The general lessons about ratings, accurate profiles and honest review requests still apply, but the percentages are not ours to carry over.
- Is Cheers a neutral source?
- Not entirely. Cheers sells AI visibility tools and services, which is why we treat its numbers as one company's measurements and read its own caveats closely. It says every finding is an association in its sample, that business names were extracted automatically, and that its counts are lower bounds. We also sell local SEO services, so weigh our reading with that in mind.
- Can I pay or reward customers to get more five-star reviews?
- No. Google's Maps user generated content policy says merchants may not offer incentives such as payment, discounts, free goods or services in exchange for a review. It also says merchants may not discourage negative reviews or selectively solicit positive ones, and lists staff being told to collect a set number of reviews as prohibited. Asking every customer, the same way, for an honest review is what the policy allows.
- Do I need special schema or an AI file to get into AI Mode?
- Google says no. Its guide to generative AI features says structured data is not required for generative AI search, that there is no special schema.org markup to add, and that you do not need to create new machine readable files, AI text files or Markdown. Google does say tools like Google Business Profiles can help your business be visible in AI responses, so keep that profile accurate.
- How many new reviews does it take to lift a 4.2 rating?
- That is arithmetic, not a Google rule. If a business has 100 reviews averaging 4.2, it needs about 60 new five-star reviews to reach a 4.5 average. Every new review that is not five stars makes the climb longer. That is why fixing the service problem behind a low rating usually beats chasing volume.
Sources & Further Reading
- Cheers, “AI search for local businesses, measured: 9,435 Google AI Mode answers” (published May 19, 2026, updated September 28, 2026): cheers.tech (read November 20, 2026): every statistic in this post, the methodology and the caveats. Cheers sells AI visibility tools. We read the full article.
- Google Search Central, “Optimizing your website for generative AI features on Google Search”: developers.google.com (read November 20, 2026): the core ranking systems line, the AEO and GEO note, the structured data and machine readable file guidance, and the Business Profile mention. We read most of the page.
- Google Maps User Generated Content Policy Help, “Prohibited & restricted content”: support.google.com (read November 20, 2026): the rules on incentives, selective solicitation, staff review quotas and unusual review patterns. The page loaded partially, and we read the review sections.
- Blacksheep Creative, “AI Search Optimization Is Just Local SEO in Disguise” (November 9, 2026): blaksheepcreative.com (read November 20, 2026): the critique of AI search sales pitches and a second description of the Cheers study. We read the first half of the article.
