AI answers lean towards the city centre. In an AIGROW Labs study of 200 random local pages, a business within 2 km of the centre was 2.41 times as likely to be named by ChatGPT, Claude or Gemini as one of the same trade farther out. The cause is open, so the study gives no reason to move.
Run a local business? AI Local Fix sets up your listings, Google profile and markup, then counts the answers that name you.
- On 200 random local pages in 20 countries, a business within 2 km of the city centre was 2.41 times as likely to be named by ChatGPT, Claude or Gemini as one of the same trade around the same city farther out.
- Of the 863 businesses the three assistants named, 57.0% lay within 2 km of the centre, against 40.7% expected from the map. The share named fell with every step out: 8.5% under 1 km, 2.6% at 5 km or more.
- The pull was clear in 8 of the 22 kinds of business, among them restaurants, hair salons, gyms and dentists. Real estate agencies showed none, and on small maps and in the cities read to 10 km the study cannot tell.
- The study shows a pattern, not its cause. Central businesses may simply be better known, so the study gives no reason to move.
- What an owner farther out controls is how clearly the web places the business: its neighbourhood, the areas it serves and the same facts everywhere. The free checker is the first step.
How far the answers lean towards the centre
You run a gym, a hair salon or a driving school some way out of town. A customer asks ChatGPT for the best one in your city, and the names that come back belong to businesses in the centre. You may be closer to that customer than any of them.
AIGROW Labs measured how often this happens. It read answers that ChatGPT, Claude and Gemini had already given for the November 2026 edition of the AIGROW AI Visibility Index: 200 local Index pages drawn at random, 10 in each of 20 countries, covering 22 kinds of business in 88 cities, with 7,198 answers. A page is one kind of business in one city.
Each business an assistant named was found on OpenStreetMap, the open map drawn by volunteers, and its distance from the city centre was measured. It was then compared with every business of the same trade around the same city: 18,198 businesses on the map, 6,921 of them within 2 km of the centre.
The 95% interval runs from 1.96 to 2.94. Each assistant on its own showed the same pull: ChatGPT 2.54, Claude 2.66 and Gemini 2.71. Comparing only businesses alike in chain status, own website and number of sites, the ratio stays at 2.35 (1.90 to 2.96).
125 of the 200 pages have a named business found on the map; the others add nothing to either side.
Seen from the other side, 57.0% of the 863 businesses the three assistants named lay within 2 km of the centre. Had each page's named businesses been drawn from its map with no regard to distance, 40.7% would have: 1.40 times the expected share, with an interval from 1.30 to 1.50.
The questions named the city only, never a district or a street, and the assistants answered from memory: no web search, and nothing told them where the person asking was. The study tested only that city-wide question.
Every step away from the centre counts
The study split each city's map into bands by distance from the centre and measured the share of businesses in each band that at least one assistant named. Each band is compared with businesses 5 km or more from the centre on the same page.
| Distance from the centre | Businesses on the map | Named by any assistant | Share named | Against 5 km and beyond, same page |
|---|---|---|---|---|
| 0 to 1 km | 3,354 | 286 | 8.5% | 5.05 (3.26 to 8.25) |
| 1 to 2 km | 3,567 | 206 | 5.8% | 2.84 (1.90 to 4.47) |
| 2 to 3 km | 2,770 | 112 | 4.0% | 2.36 (1.57 to 3.55) |
| 3 to 5 km | 3,645 | 131 | 3.6% | 1.56 (1.12 to 2.17) |
| 5 km to the edge | 4,862 | 128 | 2.6% | reference |
The share falls with every step. A business under 1 km from the centre was 5.05 times as likely to be named as one of the same trade and city 5 km or more out. Between 3 and 5 km out the gap is smaller, 1.56 times, and its interval stays above 1. The last column gives each 95% interval in brackets.
Even near the centre most businesses go unnamed: 8.5% is fewer than one in ten. An answer names a handful of businesses, so in a trade with many of them on the map most are left out wherever they are.
Where the pull is clear, and where it is not
In 8 of the 22 kinds of business, a central business was more likely to be named even at the low end of the 95% interval. The table shows those 8 and real estate agencies, most named first.
| Kind of business | Named businesses | Central share, named | Central share expected | Central vs farther out, same page |
|---|---|---|---|---|
| Restaurants | 223 | 71.3% | 46.8% | 3.16 (2.28 to 4.63) |
| Hotels | 189 | 70.4% | 54.3% | 2.69 (1.61 to 7.32) |
| Hair salons | 62 | 59.7% | 33.3% | 3.46 (2.00 to 6.90) |
| Gyms | 56 | 48.2% | 31.6% | 2.20 (1.32 to 3.12) |
| Real estate agencies | 55 | 43.6% | 47.3% | 0.83 (0.52 to 1.32) |
| Dentists | 54 | 38.9% | 23.8% | 2.25 (1.33 to 3.30) |
| Driving schools | 34 | 44.1% | 31.1% | 2.18 (1.15 to 3.47) |
| Architects | 24 | 66.7% | 54.0% | 1.79 (1.07 to 5.14) |
| Insurance brokers | 16 | 68.8% | 48.4% | 2.36 (1.02 to 3.77) |
Each kind rests on 9 or 10 pages, so read a single row for its direction only. Real estate agencies, with 55 named businesses, show no clear pull. The remaining kinds show no clear pull either, and in several of them so few businesses were named that there is little to compare.
If you run a hotel or a restaurant
Hotels and restaurants are the trades visitors from elsewhere ask about, and a visitor may well want the centre. Without hotel and restaurant pages, a central business was still 1.98 times as likely to be named, with an interval from 1.53 to 2.51, on the other 182 pages and their 7,960 businesses. This check was added after the first results and was not part of the method fixed in advance.
Big cities and small maps cannot tell
In the cities whose maps were read to 10 km, among them London, New York and Paris, the same-page ratio is 1.33 (0.90 to 3.43) on 17 pages. On the 103 pages with 15 businesses or fewer on the map it is 1.57 (0.96 to 2.28). Both intervals reach 1, so for these pages this sample cannot tell either way.
The pull is clear where a city and trade have many businesses on the map.
So if your business is in a small town, or in a trade with few businesses around your city, the study cannot say whether the pull reaches you. In a large city, keep in mind what the centre is: the reference coordinate AIGROW Labs uses for the city, not a measured downtown. In a city with several centres, or one whose businesses cluster away from that point, distance from it means less, and that may be why the cities read to 10 km show a weaker result.
Why it happens is still open
The study shows a pattern, not its cause. Central businesses may be older, better known, more reviewed or more written about, and any of these could be why the models name them. How well known a business is may explain the pattern as much as distance does, and the study cannot separate the two.
So it gives no reason to move. It does not show that location causes the gap, and a new address would not make a business older or better known. What an owner farther out can do is describe the business clearly enough for an assistant to place it, and work on what makes a business known wherever it stands.
What an owner farther out can do
The study did not test what makes an assistant name a business farther out, so the steps below are general practice and not its findings. A model answers from what it learned from the web, and assistants with search also read pages at the moment they answer. The facts an owner controls sit on their own website, their Google Business Profile, the directories for their trade and the reviews customers leave.
Start with the first step, because it shows where you stand before you change anything.
1. See what the assistants say now
Run the free checker. It puts five questions a customer in your trade and area would ask to ChatGPT, Claude, Gemini and Grok, and shows every answer word for word and who gets named instead.
Then ask an assistant the city-wide question the study used, such as the best hair salon in your city, a few times, and the same question with your district or town in it. Write down who is named and roughly where each one is. That list is your starting point, and the one to compare against later.
2. Say where you are and where you work
Name your neighbourhood and the areas you serve on your own website and on your Google Business Profile, so an assistant reading about you can place you in the city. Put them in plain words on your home page and contact page, not only in an embedded map: a family-run gym in the north of the city, a short drive from the towns along the ring road. If you go out to customers, set the areas you serve on your Google Business Profile.
3. Write for the districts you serve
Give each district or nearby town you serve a page of its own that says what you do there and how customers reach you: the bus or the parking, the work you have done there, the questions people there ask. Write only for places you really serve, and make each page say something the others do not. A page copied with the town name swapped tells an assistant nothing new.
4. Be the same business everywhere
Use one name, address and phone number on your site, your Google Business Profile, Bing Places, Apple Business Connect and the directories of your trade, so every mention an assistant reads points to the same place. The free NAP consistency checker compares them for you.
Check your business on OpenStreetMap too, with its real name, address and trade. The study looked for the named businesses there, and names it could not tie to one business on the map were left out of every count. Adding or correcting your own business costs nothing, and it is one more place where your facts are written the same way.
5. Build what makes a business known
If central businesses are named because they are better known, and the study cannot tell, then what makes a business known is open to an owner anywhere: reviews from real customers, a complete Google Business Profile, listings in the directories and associations of your trade and your town, and mentions in local news and guides. Ask every customer for a review, never offer anything in return, and never ask only the happy ones.
6. Check again on the same questions
Assistants rarely give the same list twice, so one answer is a sample. Ask the same questions a few times before and after each change, and count how often you are named. Why ChatGPT recommended your business last month and not now explains why a single answer tells you little.
Where AI Local Fix fits
The steps above take time, and most owners have little of it. AI Local Fix does the groundwork once, for one local business.
It claims, corrects or creates up to 15 listings on the directories assistants cite for your trade and country, with the same name, address and phone everywhere. It completes your Google Business Profile (category, services, hours and description) and puts the same facts on Bing Places and Apple Business Connect. It adds LocalBusiness markup to up to 10 of your pages, writes 3 answer pages from the questions your customers ask, and prepares a reviews kit.
It also measures. Before the work starts, at day 30 and at day 90, it asks ChatGPT, Claude, Gemini and Grok 10 questions for your city that you approve, and counts how often they name you next to your 3 closest rivals. For a business outside the centre, those rivals can be the central businesses the assistants name in your place.
If the day 90 count names you no more often than the first, you can ask for a full refund within 14 days of the day 90 report. Nobody can promise a ranking in AI answers, and AI Local Fix does not.
What the study cannot tell you
- It shows a pattern, not its cause. Central businesses may be older, better known, more reviewed or more written about, and the study cannot separate how well known a business is from where it stands.
- It gives no reason to move, since it does not show that location causes the gap.
- The assistants had no web search and no location. An assistant that knows where the person is, or that searches the web, may answer differently.
- The centre is the coordinate AIGROW Labs uses for each city, not a measured downtown. In a city with several centres, distance from it means less.
- The check without hotel and restaurant pages was added after the first results and was not part of the method fixed in advance.
- Big cities and small maps cannot tell: in the cities read to 10 km, and on pages with 15 businesses or fewer on the map, both intervals reach 1.
- Names that could not be tied to one business on the map are left out of every count. If they were mostly far from the centre, the figures here would be overstated; where they are was not measured.
- 200 pages, each kind of business on 9 or 10 of them, all from one edition of the Index, November 2026. There is no second sample to repeat the test on.
- AIGROW sells services that touch these findings, AI Local Fix among them. The study is a working paper, not peer reviewed, and its method, tables and limits are on the study page.
Rather have this done for you?
AI Local Fix does the groundwork once: your Google Business Profile, up to 15 listings, the markup on your pages and a reviews kit. It counts how often four assistants name you on 10 questions for your city before the work, at day 30 and at day 90.
If the day 90 count names you no more often than the first, you get the full price back. Ask within 14 days of the day 90 report. No ranking and no specific AI answer is promised.
Frequently Asked Questions
