Research
Jacksonville Business Listings Study: Method and Data
How we checked 600 Jacksonville-area businesses' directory listings against their own Google Business Profiles, what our team checked by hand, and the business-level data, with no business named. The findings are in Business Listing Accuracy: 600 Jacksonville Businesses Checked.
What did we measure?
We drew 60 businesses at random in each of ten industries (dentists, HVAC, med spas, moving, personal injury law, pest control, plumbing, pool service, real estate and roofing) from the businesses Google Maps shows for each industry's two biggest buying searches, run from 29 fixed points across Duval, St. Johns, Clay, Nassau and Baker counties.
The reference for each business is its Google Business Profile: the name, phone and address Google shows. The Citations Checker in our software, Narsil SEO, searched Google for each business by name and by phone, across the web and on up to 12 directories' own sites, 30 results deep, and compared every listing it found with the profile, field by field. The searches ran October 15 and 16, 2026.
A listing is consistent when its name, phone and address agree with the profile. A location tag in the name or a missing suite number doesn't make it wrong. A different phone number does, even a call-tracking number set up on purpose.
What did people check by hand?
The checker reads search results, and search results leave things out, so our team checked its work in a normal browser. Where a person made a call, it replaces the checker's.
| Checked by a person | How many |
|---|---|
| Every Yelp and Facebook listing the searches found | 714 listings |
| Every page our follow-up searches and a Yellow Pages fix added | 433 pages |
| Every search our data vendor failed to answer | 643 searches |
| Listings drawn to test the checker's calls | 140 listings |
| Businesses searched from scratch on six directories | 20 businesses, 76 listings |
Before collecting, we set a 10% line for how often a person may disagree with the checker. It crossed that line, so every figure we publish uses the corrected counts.
| How the checker did | Result | 95% range |
|---|---|---|
| A person's call differed from the checker's | 11 of 95 (12%) | 7% to 20% |
| Checker said consistent; the page itself showed a different phone or address | 10 of 79 (13%) | 7% to 22% |
| Checker matched the page to the wrong business | 5 of 100 (5%) | 2% to 11% |
| Listings a person found that the checker also found | 41 of 76 (54%) | 43% to 65% |
| The same, on Yellow Pages | 2 of 15 | |
| The same, on Apple Maps | 4 of 16 |
How much did the hand checks change?
A lot. With every hand call applied, the 600 have more listings than the checker found on its own, and more of them are wrong.
| Across the 600 | Checker alone | With every hand check |
|---|---|---|
| Directories with a listing, per business | 1.72 | 2.39 |
| Of those, consistent with the profile | 1.32 | 1.69 |
| Of those, contradicting it | 0.40 | 0.66 |
| Businesses with no listing found | 30% | 16% (13% to 19%) |
| Businesses with at least one wrong listing | 27% | 42% (38% to 46%) |
Ranges in brackets are 95% ranges.
Businesses with at least one wrong listing
Businesses with no listing found
What can't the data tell you?
- Every count is a minimum. Our searches surfaced about half the listings a person found (41 of 76), and far fewer on Yellow Pages and Apple Maps.
- Yelp and Facebook rest on a person's call for every listing. BBB, MapQuest and Nextdoor mostly rest on the checker's, so they likely understate how many listings are wrong.
- Bing Places and Foursquare weren't measured, so they're absent here, not zero.
- Businesses in Google's top three map results have more consistent listings, and that's a correlation: they also differ in reviews, profile age and websites.
- It describes 600 businesses in ten industries in the Jacksonville metro in October 2026, and nothing wider.
Download the data
One row per business, 600 rows. Names, phones, addresses and every link are removed, and the rows are shuffled within each industry, so no business is named.
Download the data (CSV, 54 KB)| Column | What it holds |
|---|---|
| industry | One of the ten industries |
| gbp_claimed | Whether the business's Google Business Profile is claimed: yes, no or unknown |
| listings_found | Directories where we found this business's listing (consistent, inconsistent or cannot_tell) |
| consistent | How many of those agree with the profile on name, phone and address |
| inconsistent | How many contradict it |
| cannot_tell | Listings a person confirmed as this business's but couldn't judge, mostly Facebook pages that hide the phone and address |
| angi.com to zocdoc.com | One column per directory, 27 in all, each with at least one listing among the 600: consistent, inconsistent, cannot_tell, unconfirmed (a page that names the business but shows too few fields to judge, not counted as found), or blank (none found) |
If you use it, cite it as: Narsil Creative, Jacksonville Business Listings Study, data collected October 2026, narsilcreative.com/research/jacksonville-listings/.
Who ran it?
Narsil Creative, an AI SEO agency in Atlantic Beach, Florida, in the Jacksonville metro. We sell local SEO, and the Citations Checker in our software, Narsil SEO, collected the data, so we have an interest in what it says. That's why the method and the data are here to check. We plan to re-run the study every year with the same sample frame, seeds and rules, and publish what changed.