The engine told us its criteria
In one of our personal-injury panel runs, ChatGPT opened its recommendation list by stating its method: candidates chosen on “board certification, peer recognition, relevant experience, and Texas Bar records — not advertising alone.” That sentence, from the engine itself, is the whole strategic picture for legal marketing on this surface. The firms that dominate billboards and television in the same market appeared in none of our runs. Every firm ChatGPT named was board-certified.
In Nameworthy’s personal-injury panels, ChatGPT printed its own selection criteria inside the answer — board certification, peer recognition, relevant experience and state bar records, and explicitly not advertising alone.
The verification behavior is new, and it matters
Across our legal runs, ChatGPT didn’t just rank — it verified, and said so:
- Bar discipline checks, printed in the answer. “The State Bar currently reports no public disciplinary history” appeared for pick after pick, cited to the bar’s own site. A clean record has become quotable evidence; a checkable one is table stakes.
- The certification explained to the client. Answers now teach what board certification means and how rare it is, then filter the market by it.
- Deadlines flagged from primary sources. Statute-of-limitations warnings cited to the state’s statutes site, down to municipal notice deadlines. The engine behaves like a cautious referral partner, not a directory.
- Case-type routing. Different sub-specialists recommended for trucking, medical malpractice, and catastrophic-injury questions, with the follow-up offer to narrow further.
What separates the named from the unnamed
Holding a board certification was necessary in our runs but rarely sufficient. The firms ChatGPT actually named paired it with corroborated depth: peer-review directory recognition (Best Lawyers years, Super Lawyers spans, AV ratings, Avvo scores), decades of tenure, or a documented trial record. In family-law markets we’ve also measured directory-profile depth carrying strong ChatGPT presence for firms with well-maintained Avvo profiles. The pattern across both: ChatGPT trusts what third parties can confirm. A credential that lives only on your own site is an assertion; the same credential echoed by a directory and a registry is evidence.
What a firm can actually do
- Publish your certifications in liftable form — the exact certification name, the certifying body, the attorney it attaches to, with a link to the registry where it can be verified.
- Build the peer-directory layer. The directories ChatGPT cited in our runs are specific and few. Complete, current profiles there are corroboration; absence there is invisibility on this surface.
- Keep the bar record clean and claimable — and make sure your attorneys’ bar profiles are complete, since the engine reads them.
- Publish case-type depth. Sub-specialty routing is real: pages that document your trial record and case focus in concrete terms give the engine the “relevant experience” leg of its own stated criteria.
- Then measure. This surface’s answers differ sharply from Gemini’s, Perplexity’s, and Google’s AI surfaces in the same market — winning one says nothing about the others.