Why isn't ChatGPT recommending my law firm?

If ChatGPT is not naming your firm when asked for lawyers in your practice area and city, the cause is almost always one of five things: a thin Avvo or Justia profile, missing or broken LegalService schema, no published case results, generic practice-area pages with no procedural depth, or NAP inconsistencies across the legal directory stack. ChatGPT does not have a hidden lawyer index. It reads the same web everyone else does.

How ChatGPT actually picks the firms it names

ChatGPT with browsing fetches live search results when asked a current question, then synthesises an answer using its training data plus what it just retrieved. ChatGPT without browsing falls back to training data alone. For "best [practice area] lawyer in [city]" queries, the inputs the model is most likely working from are:

• Live search results for the query, which are themselves shaped by Google's local pack and organic results.
• Indexed Avvo, Justia, FindLaw, Martindale-Hubbell, Lawyers.com, Super Lawyers, and Best Lawyers profile pages.
• State bar association directories.
• The firm's own website, if it has been indexed and the practice-area pages have substantive content.
• News mentions and substantive review text from Google reviews and Avvo client reviews.

If your firm is invisible across these inputs, ChatGPT cannot recommend you. There is no separate "AI ranking algorithm" to optimise for. The fix is making sure the firm exists, in depth, on the surfaces ChatGPT is actually reading.

The five gaps we see almost every time

When we audit a firm that is missing from ChatGPT and the other engines, the diagnosis usually comes down to some combination of:

1. Thin directory profiles. Avvo profile with no photo, partial bio, no peer endorsements, fewer than ten reviews, or primary practice area mis-set. Justia profile minimally filled. Super Lawyers and Best Lawyers absent. The directories are the citation backbone for legal queries; an empty profile is invisible.
2. Missing or invalid LegalService and Attorney schema. No JSON-LD on the homepage or attorney bios. Or schema present but with broken sameAs links, missing geo coordinates, or wrong @type. Without schema, the engines have to guess what entity the page represents - and they often guess wrong or skip the page.
3. No case results published on the firm site. Within ABA Model Rule 7.1 and the relevant state's analogue, you can publish verdicts, settlements where disclosable, motions won, and charges dismissed with appropriate disclaimers. Firms that publish substantive case results get cited as evidence of capability. Firms that do not are indistinguishable from every other firm on the local pack.
4. Generic practice-area pages with no procedural depth. A 350-word page titled "Personal Injury Services" that lists "car accidents, slip and fall, wrongful death" and a contact form is not citable content. ChatGPT lifts from pages that explain the actual legal procedure, statutes of limitation, comparative fault rules, and what the client should expect from the process.
5. NAP inconsistencies. The firm name shows up as "Smith Law" on the website, "Smith Law PLLC" on Avvo, "Smith Law Firm" on Justia, with three different suite numbers across the directories. The engines either pick one entity arbitrarily or merge them into something that looks suspicious. NAP hygiene is unglamorous and high-impact.

What is not the cause

Three things firms commonly blame that are usually not the actual problem:

• "We don't have enough blog posts." Volume is rarely the issue. Substance is. One deep state-specific guide outperforms twenty generic explainers.
• "ChatGPT just hates our firm." ChatGPT does not have firm-level preferences. It surfaces what its training and retrieval pipeline read. If competitors are showing up and you are not, the gap is in inputs, not in the engine.
• "We need to pay for AI optimisation." There is no ChatGPT-specific tactic in 2026 that is not just SEO and directory hygiene. "AI optimisation" packages that promise direct ChatGPT placement are repackaging legal SEO basics.

How to diagnose your specific gap

The fastest way to find your firm's gap: run the actual prompts a client would run, in all five engines, and read the citations the engines produce. Note which directories show up, which competitor firms get named, and what content the engines lift. The gap is usually obvious within ten minutes of looking at the output side by side.

Things to compare your firm against:

1. Search your name plus city in ChatGPT and Perplexity. Do they describe you accurately? Do they cite your site or only third parties?
2. Search your top practice area plus city. Do you appear at all? If not, which firms do, and what do they have on Avvo and Justia that you do not?
3. Pull your Avvo, Justia, FindLaw, Super Lawyers, and Best Lawyers profiles. Score them on completeness against the firms that are getting cited.
4. Check the schema on your homepage and a few attorney bios. Validate the JSON-LD. Confirm the sameAs links resolve.
5. Compare three of your practice-area pages against the highest-cited competitor's. Word count is a weak proxy. The right question is whether your page actually walks through the legal procedure or just promotes the firm.

Most firms find the answer in step 3 or 4. The fix is not glamorous. It is profile depth, schema correctness, and content with real procedural substance. That is what gets a firm into ChatGPT's recommendation set.