AI names a store.
Is it the right one?

People ask AI tools local questions in plain language: 'who does a good lawn mow in Woy Woy', 'best kids tutoring near Castle Hill', 'is the Kotara one any good'. Google says AI Overviews and AI Mode can include information about local businesses in their answers.

For a single business the question is simple: are you named? For a franchise network it has three parts. Is the brand named, is it the right location, and are the hours, phone number and services in the answer the ones that store actually has?

The answer is built from what the tool can find. Google's guide says its Search doesn't use llms.txt files, and no special schema is needed for its AI features. The facts that drive local results do the work here too.

Google calls it query fan-out: the model generates related searches to gather more information before it answers. A question about the best pest control in Newcastle may also look for prices, reviews, opening hours and whether a business covers a particular suburb.

For a network, each of those searches is a chance for a different store's page or profile to turn up. If the Charlestown profile has old hours and the Newcastle page lists services that store doesn't offer, the answer can mix them up.

Google also warns against the obvious overreaction. Its guide says fan-out shouldn't be used to create large numbers of pages for every variation, because that can breach its scaled content abuse policy. One strong page per real location beats a page for every suburb you'd like to appear in.

When an AI answer names the wrong store, the cause is usually in the facts, not the AI. These are the usual suspects, roughly in the order I'd look.

• Profiles for stores that have closed, still live and not marked closed.
• Two profiles for one store, one made by the franchisee and one by head office. Google says don't create more than one page for each location.
• Suburbs added to profile names, against Google's naming rule, so one store looks like it belongs to another area.
• Service areas stretched wider than about 2 hours of driving time, overlapping the next franchisee's territory.
• Location pages copied from one template, so every store claims the same services.
• A central call centre number on every profile, so nothing ties a phone number to a place.

Pick one source of truth for each location: name, address, phone, hours, services and service area. Head office holds it. Franchisees tell head office when something changes. Then the Business Profile, the location page and the main directories are all updated from that one list.

Google's local ranking rewards the same habit. It says businesses with complete and accurate info are more likely to show up in local search results, and that hours, special hours, reviews and photos all help. Those are the same facts an AI answer about a store needs.

Then check the parts only the store knows. Does it really take weekend jobs? Does it cover the suburb next door? The location page should say what that store does, in the words its customers use.

Enter your brand, the service people ask for and the suburbs where you have stores. Run the prompts in ChatGPT, Gemini and Google AI Mode, logged out if you can, and record which business each answer named. Answers change from day to day, so treat one round as a snapshot.

Work from the results. A suburb where a competitor gets named is a different problem from one where your own other store gets named, and they need different fixes.

• Wrong store named: check that suburb's profile, page and service area for overlaps and stale details.
• Closed store named: mark the profile closed, take it out of the store finder, and redirect its page to the nearest open location.
• Competitor named: compare reviews, profile completeness and the location page with theirs. That's a local SEO job, covered on the multi-location SEO page.
• Nobody named: the answer probably wanted something no page says, such as price ranges or weekend availability. Add it to the location page.

Google now points site owners to a Search Console generative AI performance report to see how content performs in its AI features. That shows pages, not stores, so for a network the practical measures are still calls, bookings and visits per location, plus a monthly round of the naming test above.

Keep the test small and repeatable. The same prompts, the same suburbs, roughly the same day each month. A change in which store gets named tells you more than any single answer does.