Isoline Atlas No. 01
Chapters ▾
Get my free map
Isoline Growth / Research
Isoline Atlas · No. 01 · Charlotte

The Three‑Mile Map

How far 138 Charlotte-area home service companies reach on Google Maps, and which ones AI assistants recommend.

By Neil, founder, Isoline Growth · October 16, 2026

7,056 Maps searches141,120 ranked results159 AI checks2020 Census weights
Read the map in 60 seconds
70% → 1%

Residents who see a typical company in Google's top 3: 70% near its shop, about 1% at 6.7 miles.

3%

Residents within 10 miles, outside its own neighborhood, for whom the typical company ranks top 3.

51% vs 21%

Top-3 rate 3.3 miles out, in quieter versus busier areas. Busier areas are harder to win.

27 of 53

Companies ChatGPT named at least once. Gemini named 31, Google's AI Overviews 2.

Summary

Thirteen findings

Counted by where people live, most companies in this study are visible on Google Maps only in their own neighborhood. AI assistants follow different rules, and each leaves many companies out.

70% → 1%The three-mile rule.76% of companies rank top 3 at their own address. Counted by residents: 70% near the shop, 25% at 3.3 miles, about 1% at 6.7.
Plate 3 →
3%A sliver of the market.Outside its own neighborhood, the typical company ranks top 3 for about 3% of residents within 10 miles.
Plate 3 →
51% vs 21%Busier areas are harder to win.3.3 miles out, companies made the top 3 in 51% of less-populated areas but 21% of more-populated ones.
Plate 4 →
AI assistants · findings 10–13
Chapter I

The map

Google shows most companies near their own shop and almost nowhere else.

1.1 One company's map1.2 The three-mile rule1.3 Busier is harder1.4 Each trade's cliff
70% →
1%
1.1

Visibility isn't a number. It's a map.

When a homeowner in Matthews searches for AC repair, Google doesn't show the best AC company in Charlotte. It shows well-matched companies near where that homeowner is. Google lists relevance, distance and prominence as the three factors behind local results.

So visibility changes block by block. Most owners only see one point on that map, the search from their own office, which is usually where they look best. We measured the whole map for 138 companies, weighted it by where people live, and asked three AI assistants the same question.

Plate 1#plate-1

One real company: #1 at its shop, invisible a few miles out

Where one company ranked for "ac repair" at 49 points around its shop. Real scan, name removed.

Top 3 at 1 of 49 pointsIts own shop, where it ranks #1.
Ranks 4–10 at 6 points, 11–20 at 4Visible only if the homeowner scrolls.
Outside the top 20 at 38 pointsEach small grey dot is a spot where it isn’t in Google’s top 20 for “ac repair.”
Top 34–1011–20Not in top 20
Each dot = one spot we searched from, about 3.3 miles apart · Rank data collected with Local Falcon · Map data © OpenStreetMap contributorsIsoline Growth · The Three-Mile Map
Plate 2#plate-2

The average map: 76% at the shop, single digits two rings out

Share of all 144 company scans ranking top 3 at each grid point, with the shop at the center. Darker squares mean more companies won that spot.

The shop: 76%Three in four companies rank top 3 at their own address.
First ring, 3.3 mi: 22–41%Already well under half, on every side.
Second ring, 6.7–9.4 mi: 2–8%Single digits at every point.
Edge, 10–14 mi: 0–3%Most companies are invisible here.
Unweighted grid points, 144 company scans · Rank data collected with Local Falcon, Oct 15, 2026Isoline Growth · The Three-Mile Map
1.2

The three-mile rule

Counted by where people live, companies were top 3 for one in four residents 3.3 miles away, and about one in a hundred at 6.7 miles.

Plate 3#plate-3

Counted by residents, visibility falls faster than the grid suggests

Share in Google's top 3 by distance from the shop. Shaded band: 95% range from resampling companies.

Residents (population-weighted)Grid points (unweighted)
144 company scans, 7,056 searches · residents = 2020 Census block-group population in each 3.3-mile cell · 2,000 company-level bootstrap samplesIsoline Growth · The Three-Mile Map
¼
Not even in the list.

At 6.7 miles, the company appeared anywhere in the top 20 for about a quarter of residents.

3%
The typical reach.

Within 10 miles, outside its own cell, the median company ranks top 3 for 3% of residents. The average is about 10%, pulled up by a few far-reaching companies.

31 of 144
Never top 3.

Scans with no top-3 spot anywhere on the grid, not even at the shop.

Where does your map drop off?

Get the same 49-point scan, free →
1.3

Where people live is hardest to win

At the same distance from the shop, companies were far less likely to rank in the busier parts of the map. That's why the resident curve falls faster.

Plate 4#plate-4

At 3.3 miles, companies won half the quieter areas and a fifth of the busier ones

Top-3 rate at each distance, splitting grid cells at the median population for that distance

Less-populated cellsMore-populated cells
Median cell population at 3.3 mi: about 6,600 (less) vs 22,500 (more) residents · association only; denser areas likely hold more competitorsIsoline Growth · The Three-Mile Map
1.4

Each trade has its own cliff

The drop shows up in every trade. HVAC, the most crowded, falls off first: about 0.1% of residents at 6.7 miles.

Plate 5#plate-5

Seen in cross-section, HVAC's peak is the narrowest

Resident-weighted share in Google's top 3, drawn as a profile through the shop, 14 miles each way. Median competing businesses across each company's 49 searches in brackets.

HVAC 82, pest control 26, plumbing 23, garage door 13 company scans · differences outside HVAC are within the margin of error at these sample sizesIsoline Growth · The Three-Mile Map
Chapter II

What wins

Closeness does the most. Reviews matter up to a point. No one owns the map.

2.1 Who holds each position2.2 Reviews and reach2.3 Nobody owns the map
150
2.1

What it takes to win a search

Distance shows the strongest link to winning. #1 sat 3.7 miles closer than #11–20, with three times the reviews.

Plate 6#plate-6

Each step down the results is farther away, less reviewed and lower rated

Medians across all 141,120 ranked results, by position

Distance = from the searcher's grid point · rating = average; its scale starts at 4.5Isoline Growth · The Three-Mile Map
4.5★
The rating floor.

93% of top-3 results carried a rating of 4.5 or higher.

99%
Category is an entry ticket.

Of #1 results, 99% had a primary Google category matching the trade searched.

503
Plumbing winners carry the most reviews.

Median reviews of the #1 plumbing result, about twice HVAC's 259. In garage door, 80.

2.2

Reviews get you on the map

Under 50 reviews, companies were mostly invisible. Past about 150, more reviews showed no relationship with reach.

Plate 7#plate-7

Past about 150 reviews, the line goes flat

Each dot is one company: its Google reviews (log scale) and the grid points, of 49, where it ranked top 3. Line: running median.

Won a top-3 spot somewhereNever top 3Running median
144 company scans · rank correlation 0.28 overall; −0.02 among the 93 with 150+ reviews (not significant) · association, not causeIsoline Growth · The Three-Mile Map
2.3

Nobody owns the map

Across 7,056 searches, 21,168 top-3 spots went to 1,338 different companies. The company beating you three miles away is usually another local shop.

Plate 8#plate-8

In every trade, the five biggest winners hold under a fifth of top spots

Each square is 1% of the trade's top-3 spots in our searches

Five most frequent companiesNational brands and franchisesEveryone else
Counted across our searches, so shaped by where the scanned companies sit · national brands identified by name, so this is a floor · the two groups can overlap slightlyIsoline Growth · The Three-Mile Map
<1.5%
No one dominates.

The most frequent HVAC winner's share of HVAC top-3 spots in our searches.

60.5%
Deals aren't map share.

Private equity sponsors' share of residential services M&A deals in 2026 through July 27 (Capstone Partners). Franchise and national brands held 6% of top-3 spots here.

17%
Garage door is the exception.

National brands' share of garage door top spots, in the trade with the fewest competitors.

Chapter III

The AI layer

AI assistants don't follow the map. Gemini repeats a few names everywhere; each assistant leaves many companies out.

3.1 Who gets named3.2 A few names everywhere3.3 Where the answers come from
Half left out
3.1

Who gets named

ChatGPT and Gemini each left about half the companies unnamed. Google's AI Overviews named two.

Plate 9#plate-9

ChatGPT and Gemini each named about half of 53 companies

Each square is one company we checked; filled squares were named at least once. The bracket shows the 95% range.

One check per assistant on a 3×3 grid · one company named by all three, eleven by none · AI Overviews returned sources in 16 of 53 checks; our data can't separate low coverage from companies not being chosenIsoline Growth · The Three-Mile Map
3.2

A few names, everywhere

On Maps, the five biggest HVAC winners held 7% of top spots. On Gemini, 31% of HVAC recommendations went to five companies.

Plate 10#plate-10

Gemini gave five companies a third to over half of its recommendations

Share of recommendations (or Maps top-3 spots) going to the five most frequent companies, by trade

Maps: 21,168 top-3 spots · AI: 2,308 Gemini and 1,288 ChatGPT recommendations, one run each · name variants counted separatelyIsoline Growth · The Three-Mile Map
9 of 9
Gemini's picks covered the grid.

When Gemini named a company, it named it at a median of 9 of 9 points. ChatGPT: 1 of 9.

24 of 45
Separate scoreboards.

Companies with any Maps top-3 spot that ChatGPT named. A Maps lead didn't carry over reliably.

1 run
A snapshot, not a rule.

AI answers vary between runs. These are one run per company, through simulated queries.

3.3

Where the answers come from

Three directory sites showed up in the sources of every ChatGPT check.

Plate 11#plate-11

ChatGPT's cited sources lean on directory sites

Share of ChatGPT checks whose cited sources included each site

53 ChatGPT checks, sources pooled across each check's 9 points · Gemini more often cited company websites and ForbesIsoline Growth · The Three-Mile Map
What to do

What it means for owners

A safe working assumption: Google shows you well in your own neighborhood and poorly a few miles out, especially where most of your customers live.

  1. Look at your map, not your search.Your own search shows your best spot. Check what homeowners see five and ten miles out.
  2. Compare yourself with the neighbor.The company ahead of you in Matthews is usually a local shop closer to Matthews.
  3. Hold 4.5 stars and the right category.Both are close to an entry ticket. Your primary category should name your main trade.
  4. Get past 50 reviews.In our sample, companies under 50 had a median of zero top-3 spots. Beyond about 150, count alone showed no link to reach.
  5. Pay where the map won't reach.Local Services Ads and search ads can target the zip codes the free map doesn't carry you to.
  6. Keep directory profiles accurate.Thumbtack, Angi, HomeAdvisor and the BBB showed up in ChatGPT's sources across our checks.
  7. Ask the AI yourself.Ask ChatGPT and Gemini who to call for your service in your town, and see who they name.
Take it with you

Download the full report as a PDF

All 11 plates and the method notes, formatted to print or forward to your office manager.

We'll email you when the next edition is out. No other lists. See our privacy policy.

Your map

Measure your three‑mile map

We'll run the same 49-point scan for your main money search and show where you drop out, who takes those searches, and the one change most likely to move it.

Measure minehello@isolinegrowth.com
FreeOne-page mapBack in 2 business daysNo sales call required
Method

How we measured

Companies
138

Charlotte-area HVAC, plumbing, pest control and garage door companies; 144 company-and-search pairs.

Maps grid
7 × 7

49 points, 10 miles each way from the shop, 3.3 miles apart. Top 20 results recorded at each.

Weighting
2020 Census

Each point weighted by residents in its 3.3-mile cell, from block-group population centers.

Uncertainty
2,000

Bootstrap resamples of whole companies give the 95% ranges shown.

AI sample
53

Companies checked once on ChatGPT, Gemini and Google AI Overviews, 3×3 grid, same radius.

Date
Oct 15, 2026

About 7 to 11:10am Eastern. Searches: hvac repair, plumber, pest control, garage door repair.

How this compares with other research 4 agree · 2 partly · 1 differs
Distance decides most rankings
Google lists distance among its three local ranking factors. In Whitespark's 2026 expert survey, proximity ranked #2 of 187 factors.
Agrees
Visibility drops within a few miles
Rankings.io (2025) found law firm rankings hold about a mile, then fall about 8 positions. A different industry, same shape.
Agrees
Primary category is near-mandatory
Whitespark's experts rank the primary Google category the #1 local pack factor.
Agrees
93% of top-3 results rated 4.5+
In BrightLocal's 2026 survey, 31% of consumers only use businesses rated 4.5+, up from 17%.
Agrees
Gemini's picks didn't vary with distance
Local Falcon (2025) found almost no link between distance and position in Google's AI Overviews, a different product from Gemini.
Partly
ChatGPT's sources lean on Thumbtack, Angi, HomeAdvisor
BrightLocal (2026), across 1.9 million AI citations, found Google Business Profile and Yelp cited most. Our check counts presence; theirs counts share.
Differs
AI Overviews named 2 of 53 companies
Whitespark saw AI Overviews on about 15% of local-intent searches in its 2025 sample; Local Falcon reported 40% across its own sample.
Partly
Limitations 6 notes
  • Not a random sample of Charlotte. The 138 companies come from our research list, lean toward independent shops, and 57% of company scans are HVAC. The 2,933 other businesses in their results were not chosen by us.
  • One day, one search term per company. Rankings change daily and by wording ("AC repair" may differ from "hvac repair").
  • Residents, not searches. Population weighting approximates where demand is. It doesn't measure actual search volume.
  • Associations, not causes. Reviews, density and reach move together here; other factors, such as competition and location, may explain the patterns.
  • AI checks are a snapshot. One run per company through Local Falcon's simulated queries, which may differ from the consumer apps. Name matching between AI and Maps results is approximate.
  • Overlapping grids. Nearby companies' grids overlap, so busy areas count more than once in the fragmentation figures.
Next editions Raleigh–Durham, Upstate SC, Columbia

The same method runs in any metro. Planned: Raleigh–Durham, Greenville–Spartanburg and Columbia, then a Carolinas comparison. Future editions will add a second search term per trade and repeat the AI checks three times.

Sources 10