Car Dealership AI Visibility Case Study | Rank Prompt

AI knew Arlington. It didn’t know Fort Worth.

Owings Auto runs two lots fourteen miles apart. AI assistants recommended one of them constantly and the other almost never. The weekly reports showed why, and 141 days later the weaker market led.

Visibility by location, every week

Share of tracked AI answers naming Owings Auto, 9 April to 28 August 2026

Fort Worth Arlington
54.2% 49.2%

0%
15%
30%
45%
60%

1 2 3 4

Fort Worth 54.2%

Arlington 49.2%

Fort Worth

Arlington

Apr 9 May 7 Jun 5 Jul 3 Jul 31 Aug 28

  1. Fort Worth guide published
  2. Two market audit
  3. Location page rebuilt
  4. Reddit comments live

Overview

How a Fort Worth car dealership got recommended by ChatGPT

Owings Auto is a family run Buy Here Pay Here dealership with 60,000 customers behind it and its own service shop out back. Arlington is home. Fort Worth is fourteen miles up the road.

The two lots share inventory, a financing desk and a website. Ask ChatGPT or Gemini for a Buy Here Pay Here lot in Arlington and Owings came up in about a third of the answers. Ask about Fort Worth and it came up in 13% of them, behind three dealers who had owned that answer for years.

Over the first six weekly reports Fort Worth averaged 15.3%. Arlington averaged 36.3%. Owings sat twelfth among 247 businesses named in Fort Worth answers, and three of the eight tracked answers never named the dealership at all.

It was not a content problem

Running the two markets head to head in June ruled out the obvious answer. The Fort Worth blog was already cited more often than the Arlington one. What Fort Worth lacked was corroboration from other sites, and a location page worth citing.

Same brand, measured side by side

Signal Arlington Fort Worth
Rank in the local field #1 of 240 #4 of 225
Homepage citations 46 times 14 times
Outside profiles citing Owings 9 sites 2

The number that set the plan

In Fort Worth answers, location and area pages were the second most cited page type at 21% of every citation, ahead of financing pages, blogs and directories.

Competitors were winning Fort Worth with their location pages. Owings had one that read like a brochure.

The work

Four moves, in order

01

Measured each lot separately

Most tools report one score for the brand. Rank Prompt tracks every location on its own, every week, on prompts that never change, which is the only reason the gap was visible at all.

Engines tracked

Prompt categories

Every week, unchanged, since April

15 prompts × 8 answers each × 2 locations = 240 answers a week.

02

Wrote the content the answers were reaching for

Rather than filling a calendar, every piece aimed at the three tracked categories and at the questions assistants were already answering badly.

03

Rebuilt the location page from the pages AI already cited

Location pages were 21% of Fort Worth citations, so the winners became the brief. The ten most cited pages in the market gave up their strongest structural element each, matched against what Owings has and the competition does not.

04

Commented in the Reddit threads AI actually cites

Reddit was the fourth most cited domain in Fort Worth answers. Every Reddit thread cited anywhere in the account was ranked by citation count, and the top of that list got the work.

The outcome

Fort Worth passed Arlington

In April the satellite trailed the home market by 23 points. On 28 August it led by five.

9 Apr 28 Aug
Fort Worth visibility 13.3% 54.2%
Arlington visibility 36.7% 49.2%
Rank in the Fort Worth field #12 of 247 #2 of 138
Answers naming Owings 16 of 120 65 of 120

Every category moved

Fort Worth visibility by prompt category

9 April 28 August

Their own site became the most cited source in the market

In the latest Fort Worth report, owings-auto.com is cited more often than any competitor domain and more often than Yelp. In April it was not in the top six. Brand wide citation share went from roughly 1% to 5.5%.

owings-auto.com 132 luxuryautoinc.com 103 drivecasa.com 72 cowtownusedcars.com 66 russellsmithauto.com 60 yelp.com 57

The traffic followed

Referral traffic from AI assistants to owings-auto.com, over the same window.

What's next

Second place is not the finish line

  1. Take the top spot
  2. Close the off-site gap
  3. Fix the stale pages
  4. Run the pattern again

Why this worked. Tracking each location separately, every week, on prompts that never change is what exposed a 23 point gap inside a single brand. Reading the citation data underneath it is what turned that gap into four specific jobs instead of a vague instruction to publish more.