This week the tracking engine went live. It now asks Perplexity, Google AI Overview and Gemini the questions people use when they are choosing a firm rather than researching a topic, and records who gets named and which sources each answer was built from. The counts above are read live and will keep climbing as coverage widens.
Bedford appears in most answers about its own market. Dallas, Fort Worth and Houston are at or near nothing. That is not a reflection of those offices: Fort Worth carries 852 reviews and a 4.8 rating and is not named once. What separates Bedford is that it is the office the engines can resolve as an entity, and the table below groups the markets by whether they are genuinely distinguishable from one another rather than by a bare percentage.
The mechanism is visible rather than inferred. Gemini reports the searches it runs before answering, and on Dallas personal injury it searched "super lawyers personal injury Dallas TX" and "avvo top personal injury lawyers Dallas TX" by name. Across every answer collected, roughly a third of what the engines read was a legal directory. The one time Google's AI Overview named the firm, it reached it through a Justia listing for Phillip Galyen.
Alongside that, the foundation work continues to show in the inbound data. Bing AI citations and Google AI-feature impressions are both up sharply against the pre-engagement baseline, and both movements clear the noise floor rather than sitting inside normal variation. The exact figures are at the top of this page and update themselves.
| Consistently surfacing | ||
| Bedford, TX | 39 of 53 answers | 74% |
| Texarkana, TX | 22 of 33 answers | 67% |
| Mesquite, TX | 18 of 30 answers | 60% |
| Weatherford, TX | 19 of 33 answers | 58% |
| Carrollton, TX | 16 of 32 answers | 50% |
| Surfacing sometimes | ||
| Odessa, TX | 12 of 30 answers | 40% |
| Burleson, TX | 11 of 30 answers | 37% |
| Arlington, TX | 16 of 50 answers | 32% |
| Plano, TX | 8 of 30 answers | 27% |
| Effectively absent | ||
| Springdale, AR | 2 of 26 answers | 8% |
| Dallas, TX | 4 of 63 answers | 6% |
| Fort Worth, TX | 2 of 57 answers | 4% |
| Houston, TX | 0 of 57 answers | 0% |
| El Paso, TX | 0 of 36 answers | 0% |
| Little Rock, AR | 0 of 28 answers | 0% |
The sources cited most often in answers that did not name the firm.
Widen the battery to fifteen metros and six practice areas so the practice-level picture becomes testable. At present personal injury, family law and criminal defence cannot be told apart from one another with the samples collected, so any claim about which practice performs best would be premature.
Begin directory acquisition against the ranked list, starting with Justia, Super Lawyers and Expertise, which between them account for the majority of cited sources in the answers we lose.
Bind the remaining offices into the entity graph, so each city has something an engine can resolve rather than one Bedford-shaped entity standing in for twenty-two locations.
Irfad Imtiaz · Ranql