Which legal directories do AI engines cite most, ranked by engine

Which directories do AI engines cite most for law firms?

Five directories carry most of the work: Super Lawyers, Justia, Avvo, FindLaw and Martindale-Hubbell. Citorian ran 359 lawyer prompts across ChatGPT, Perplexity, Claude, Gemini and Google AI Overviews in June 2026 and logged every source cited. Super Lawyers took 33% of citations, the largest share of any single domain. Martindale-Avvo’s April 2026 analysis of its own legal query data named Super Lawyers, Avvo, Martindale-Hubbell and FindLaw as the four sources ChatGPT cites most.

The roster splits once the engines are separated. ChatGPT retrieves through Bing and OpenAI’s own crawl and produces a directory list that does not match Google’s. Gemini and Google AI Overviews read Google, so the directory that ranks is usually the directory that gets cited. Perplexity and Claude track Google’s top ten closely but Perplexity adds heavy community weighting on top. The table below is the roster a law firm should plan against.

Why does ChatGPT cite a different directory list than Perplexity?

Because the two engines read different indexes under different licensing. Martindale-Avvo found that ChatGPT’s legal answers overlap Google’s top ten results less than 25% of the time, while Perplexity and Claude sit around 75% and Gemini around 50%. A directory that dominates Google for a practice area will therefore feed Gemini, Claude and Google AI Overviews far more reliably than it feeds ChatGPT.

Perplexity then adds a second difference. Profound data reported by Claude for Lawyers puts Reddit at roughly 46.7% of Perplexity’s top-ten source share, which is more than any legal directory holds in that engine. The same data puts Wikipedia at 47.9% of ChatGPT’s top-ten source share. Perplexity also carries a commercial tie into the category: LegalZoom announced a partnership with Perplexity on June 4, 2025 that places LegalZoom services and subscriber discounts inside Perplexity answers.

Directories have noticed and are buying distribution directly. Best Lawyers launched a ChatGPT app on April 22, 2026 that answers “find me a [specialty] lawyer in [location]” from its peer-review rankings. That moves a directory’s data inside the conversation rather than leaving it to be retrieved, which is a different and more durable position than a citation.

Do the same directories get cited for every kind of question?

No. The citation layer holds across question types but the mix inside it moves. 5WPR and Haute Lawyer tested finder, decision and elite queries across eight practice areas for the Legal AI Visibility Index published in April 2026 and found seven directories owning the citation layer throughout: Chambers, Legal 500, Super Lawyers, Best Lawyers, Martindale-Hubbell, Avvo and Justia. Zero law-focused editorial sources appeared in top results, and even Am Law 100 firms ranked below the directories for queries about their own practice areas.

Local phrasing changes the surface the citation appears on. Whitespark’s Q2 2025 study of 540 local-intent queries across Houston, Phoenix and Denver found an AI Overview on 15% of plain local queries against a local pack on 93%. Hybrid queries, the longer kind that mixes a city with a circumstance, flipped it: an AI Overview appeared 97% of the time and the local pack only 17%. In the personal injury slice of that study the AI Overview sources skewed to Super Lawyers, FindLaw, Justia, Reddit and Quora.

• •Short local queries still resolve in Google’s map pack, where Google Business Profile does the work and directories matter less.
• •Longer situational queries resolve in an AI Overview, where the directory profile and the community thread do the work instead.
• •Elite and comparison queries pull the credentialing directories, Chambers, Legal 500 and Best Lawyers, ahead of the consumer directories.
• •No engine in these studies cited a law firm’s own website as the primary source for a “who should I hire” question. The firm site is the corroboration, not the citation.

Which sources get read without getting the credit?

Community sources do most of the unseen work. Ahrefs data reported by Lexgro shows ChatGPT pulling Reddit pages in 67.8% of its retrievals while naming Reddit as a source in only 1.93% of answers. A law firm reading citation lists alone will conclude Reddit barely matters, then lose a recommendation to a thread it never knew existed. The gap between what an engine reads and what it prints is the most misread number in this category.

The practical reading is that a directory profile buys a citation and a community mention buys the underlying belief. Both are needed, and only one of them shows up in a report.

Which directory profiles should a law firm fix first?

Work in the order the citation data supports, not alphabetically. Super Lawyers first, because it carries the largest single share of citations in the Citorian sample and appears in every engine’s roster. Avvo, Martindale-Hubbell and FindLaw next, because those three plus Super Lawyers are ChatGPT’s most-cited set and ChatGPT is the engine least likely to follow Google. Justia next, because it is the strongest performer in the Google-fed engines. Then Google Business Profile, then the credentialing directories that fit the practice.

Paying for a listing does not buy a citation. A complete and identical profile is what removes the corroboration gap, and a paid tier with a stale office address is worse than a free tier that matches the firm’s site. A law firm also cannot advertise a citation as an endorsement. Outputs vary between sessions, so a claim like “recommended by ChatGPT” is not verifiable in the way ABA Model Rule 7.1 requires, and the firm’s ethics counsel decides what the firm publishes about it. Firm AEO treats a dated, reproducible observation with the prompt shown as the only defensible version.

Measurement follows the same split as the roster. Firm AEO measures appearance rate per engine across an agreed set of client questions, run logged out and repeated, because a firm can hold a strong position in Gemini and Google AI Overviews while being absent from ChatGPT entirely. A single blended number hides exactly the problem the directory work is meant to fix.