Can a local business beat a national brand in AI answers?

You run a good local business. Somewhere across town, a national chain runs a polished website with a marketing team behind it. When a customer asks ChatGPT or Google who to call, it is easy to assume the machine sides with the chain. That assumption keeps a lot of owners from even trying, and it deserves to be tested rather than believed. This post puts data against it, in plain terms, and gives you a place to start this week.

One honest note before the numbers. Our dataset is young. It is large enough to see patterns and too small to declare rules, so you will read careful language here, and every figure comes with its sample size attached.

What does it mean when an AI tool names a business?

Ask an AI tool a question a customer would ask, something like "who can service a heat pump in Ballarat", and you get a written answer, not a list of links. That answer often names two or three specific businesses. When a business appears in that answer, we say it has been named, or cited. Google has its own version of this called an AI Overview, the AI-written summary that now sits above the normal results on many searches.

Those names are the new shortlist. A customer who asks for a recommendation is usually ready to act, and many never look past the answer they are given. If the tools name your competitors and not you, you lose customers you never knew were looking. That is the stake for a local business, and it is why the question of whether big brands own this space matters so much.

Do the big brands own AI answers the way they own Google search?

The fear comes from the old game. In traditional search, the work of getting found is called SEO, search engine optimisation, and one of its long-standing ideas is domain authority, an industry score estimating how much weight a website carries. Big companies tend to hold high scores, so page one of Google can feel locked from the outside.

Here is what the companies themselves disclose: not much. Google's own guidance says there are no special requirements for appearing in its AI features beyond being findable in search, and no AI company publishes a rulebook. What the industry infers, from studies and experience, is that these answers reward content that reads like a direct, trustworthy response to the question asked. That inference keeps shifting, which is worth remembering whenever someone sells you certainty.

This is no longer a fringe audience either. In a Pew Research Center survey of 5,119 American adults this February, about half said they use AI chatbots, and six in ten said they read the AI summaries at the top of their search results. The people asking machines for recommendations are your customers, now.

What we can add is what we observe in our own testing, and it does not look like a locked door.

What did our testing find about local versus national businesses?

The GetRecommended.io AI Visibility Index tests real small business websites by asking AI tools the questions a customer would ask, then recording who gets named. We have run two rounds so far: a June 2026 round covering 74 website tests (65 classified as local or national in reach) and a smaller round of live scans through July and August (40 tests, 33 classified).

Three things stood out.

In the larger June round, local businesses held their own. The local group (43 businesses) and the national group (22 businesses) were named at essentially the same rate. Both groups averaged more than 4.5 of the 5 AI tools we tested, and among Australian businesses the local group edged slightly ahead. Being local was not the handicap the domain authority mindset predicts.

The newer round leans the other way, and we say so. In the July and August scans, the national group (17 businesses) averaged roughly 4 of the 5 tools while the local group (16 businesses) averaged closer to 3. These groups are small, the direction contradicts the larger round, and the gap is not big enough or consistent enough to call a finding. We publish it anyway, because watching a young dataset honestly matters more than a tidy story.

Across both rounds, nobody was invisible. Not one of the 98 classified businesses went completely unnamed across every tool. The difference between businesses is not in or out. It is how often they are named, and for which questions. That is a contest a local business can enter, because the entry ticket is not size.

You may have seen scarier claims than ours, including industry reports that say most local businesses are invisible to AI tools. Read the method before you panic. Measurements like that typically test large multi-location brands with broad, generic questions, and a generic question rarely surfaces any one location. We ask the specific questions a business's own customers ask, and on those questions real businesses get named again and again. Both pictures can be true at once. The lesson is not that you are invisible. It is that specific beats generic, which happens to be the one contest a local business is built to win.

The usual caution applies to all of it: these are associations, not causes. Named businesses share certain traits. That is not proof the traits produce the naming.

What separates the businesses AI tools name?

If size does not decide it, something else does. Our earlier testing, published in why some businesses never appear in Google's AI answers, found the mentioned businesses shared a pattern of public proof: a deeper base of Google reviews, and pages that answer customer questions in everyday words. Technical website health did not separate the groups, and the overlooked businesses had published more blog posts, not fewer.

Notice what is absent from that list. Marketing budget. Office count. A famous name. The separating signals are all things a local operator can build, and several of them, honest reviews from real customers and specific local knowledge, are easier for you to earn than for a head office three states away.

The Local Proof Stack: what to build first

We call the pattern the Local Proof Stack. Three layers, in order.

1. One strong Google trust signal. Your Google Business Profile, the free listing that appears in Maps and search, and the reviews attached to it, work as a single signal of a real, trusted business. Keep the profile complete and current, and build a steady rhythm of asking happy customers for a review. Steady beats spectacular. A review a week compounds.

2. Answer-shaped pages. Write pages that ask and answer the exact questions customers bring to you, in their words, not industry language. "How much does a rewire cost in an older home" earns naming in a way "our electrical services" never will. Your enquiry inbox and phone log are the source material. This layer has peer-reviewed weight behind it: a Princeton-led study presented at KDD 2024 found that changing how a page presents its evidence lifted visibility in AI answers by up to 40%, with the caveat that what works varies by field.

3. Local specificity. Name your suburbs, your service area, your trade qualifications and the jobs you handle. AI tools assemble answers to specific questions, and the business that says precisely what it does and where wins the specific ask. Vague loses to specific, even when vague is bigger. We have a fuller walkthrough of this layer in how to improve AI visibility for a local service business.

What should you do this week?

You can start the stack in an afternoon and be ahead of most competitors, national ones included, inside a month.

1. Today: run the check. Ask ChatGPT for your service in your suburb the way a customer would. Search Google the same way and read the AI Overview. Write down who gets named. If you want the step-by-step version, we keep one at how to check if your business shows up in AI search.
2. This week: claim and complete your profile. Fill every section of your Google Business Profile. Then ask your three most recent happy customers for a review, and make that ask part of finishing every job.
3. Next week: publish one answer page. Take the question customers ask most and answer it on its own page, plainly, with your area and pricing guidance where you can.
4. Within the month: re-run the check. Ask the same questions again and compare notes. Visibility in AI answers moves, and watching your own result teaches you more than any generic advice.

The bottom line

The AI answer box is not a members-only club for national brands. In our testing so far, local businesses that built public proof stood level with the big names, and every business in the dataset was findable somewhere. The door is open. The owners who walk through it early, while most of their competitors are still assuming it is locked, are the ones the machines will have already learned to recommend. Find out who gets named for your patch today, with a free AI visibility scan, and if you are wondering what the scan covers, the answers are here.