ChatGPT Visibility for Law Firms: How AI Actually Recommends Firms

Just three years ago, if someone in your city got rear-ended on the highway, their first move was searching Google for "personal injury lawyer near me." You knew exactly how to approach that.

Now, more of those people use ChatGPT and type in questions that sound like real conversations, such as: "I got hit by a truck in Tampa, the insurance company is offering me $8,000, do I need a lawyer and who should I call?"

ChatGPT answers. It names two or three firms. Your firm may or may not be one of them, and there is no dashboard anywhere that tells you when you were left out.

That is the problem this guide solves. Not with hype, and not with the recycled advice you have probably already read. Below is what the evidence actually shows about ChatGPT visibility for law firms, what genuinely influences whether your firm gets named, what is being sold to you that does not work, and the order in which to fix things.

How ChatGPT actually chooses which law firm to recommend

Let go of the idea that there is a ChatGPT ranking algorithm with a position one and a position ten. That mental model will lead you to waste money.
ChatGPT is not a search engine that ranks pages. It is a language model wrapped around a retrieval system. When someone asks it for a lawyer, roughly this happens:

1. It decides whether to search at all. For "what is comparative negligence," it may answer from memory. For "best DUI attorney in Sacramento," it almost always runs a live web search, because it knows the answer changes and it needs current sources.

2. It retrieves candidate sources. OpenAI's own documentation confirms it runs a dedicated search crawler, and OpenAI has acknowledged that it may also obtain URLs from a third-party search provider. Microsoft's Bing index has been the most widely documented of those providers since ChatGPT search launched.

3. It reranks and filters. This is the step that ruins the "just rank number one" theory. Retrieval order and citation order are not the same thing. The model is selecting sources it can extract clear, corroborated, specific facts from.

4. It writes an answer and cites sources. It is summarizing what it just read. If your firm's credibility is not legible on the pages it read, you do not get named, no matter how good your firm actually is.

Three practical consequences follow from this, and they matter more than any tactic in this article:

Being retrievable is a hard prerequisite. If the crawler cannot reach you and no index contains you, nothing else you do matters.

Third-party sources often decide your fate, not your own website. ChatGPT is reading what the internet says about you as much as what you say about yourself.

The answers are not stable. Ask the same question three times, and you can get three different lists. One screenshot where you are missing is not a diagnosis, and one screenshot where you appear is not a victory.

Why ChatGPT is not recommending your firm?

In practice, the reason may be one of the following eight. Work through them in order.

1. You blocked the wrong bot, or your firewall did it for you

This is the single most common technical cause, and it is usually an accident from 2023.

OpenAI runs several separately named crawlers, and its documentation is explicit that each one is controlled independently:

OpenAI's wording on the search crawler is direct: sites that opt out of OAI-SearchBot will not be shown in ChatGPT search answers. A great many law firm websites blocked "AI bots" wholesale during the 2023 scraping panic and never revisited it. Others are blocked at the security layer, where a WAF or bot-mitigation rule serves the crawler a 403 or a rate-limit error while robots.txt says "allow."

What to do: Open yourdomain.com/robots.txt right now and read it. Then ask your developer to pull server logs and confirm OAI-SearchBot is actually fetching pages and receiving 200 responses. OpenAI notes it can take roughly 24 hours after a robots.txt change for its systems to adjust.

2. Your site is not in the index the retrieval layer draws from

Most law firms have never once checked their presence in Bing. They set up Google Search Console in 2019 and stopped there.

Set up Bing Webmaster Tools, import your Google Search Console profile, submit your sitemap, and check how many of your practice area and location pages are actually indexed. Enable IndexNow so new and updated pages are discovered in hours instead of weeks. Microsoft states plainly that IndexNow helps ensure AI systems reference the most current version of a page.

While you are there, claim Bing Places for Business. Microsoft specifically recommends it so that address, hours, and contact details stay current and eligible for inclusion in AI-generated responses.

3. Your firm is thin where AI actually looks

This is where most firm owners are surprised.

BrightLocal's August 2026 analysis ran nearly two million citations across 1,355 business locations and broke the results down by industry. For the lawyer category, the most-cited sources were Google Business Profile (42% of citations in that vertical), Yelp (7.9%), Justia (3.2%), Avvo (2.5%), Lawyers.com (1.8%), Facebook (1.7%), FindLaw (0.9%), and the Better Business Bureau (0.5%).

Read that list again. Almost none of it is your website.

Two honest caveats. That vertical breakdown covers a small number of law firm locations inside a much larger study, so treat the exact percentages as direction rather than gospel. And those figures combine Google AI Overviews, Google AI Mode, and ChatGPT, which behave differently from one another.

The ChatGPT-specific finding in the same study is the striking one: Yelp appeared as a source in 80% of all ChatGPT local answers, and ChatGPT drew on far fewer domains than Google's AI surfaces did, roughly 6,025 unique third-party domains compared with 81,817 for AI Mode.

That narrow source pool means your absence from any one major platform hurts more in ChatGPT than it does in Google's AI results.

4. Yelp just became much more important than you thought

On July 23, 2026, Yelp and OpenAI confirmed a data licensing deal that brings Yelp reviews, ratings, photos, and business details directly into ChatGPT's local answers, with Yelp branding and links attached. Yelp's "Request a Quote" feature is coming to ChatGPT as well, which means a prospect will be able to request a consultation from a service provider without leaving the chat window.

Most US law firms treat Yelp as a nuisance platform they gave up on years ago. That was defensible in 2022. It is not defensible now. A neglected Yelp profile with three old reviews and no photos is now a gap in the answer itself.

5. Your credibility is asserted but not evidenced

Look at your practice area pages the way a retrieval system does. It is trying to extract specific, checkable facts.

"We are aggressive advocates who fight for maximum compensation" contains zero extractable facts. It could describe any of 40,000 firms.
Compare that to a page that clearly states the firm handles construction site injury cases in Cook County, that the lead attorney has tried 31 cases to verdict, that the firm recovered a specific verdict in a specific type of case in a specific year, that consultations are free and fees are contingent, and that the office is open at a named address with named intake hours.

That second page gives a model something to say about you. The first gives it nothing, so it moves on to a firm that made the job easier.

6. Your reviews are stale, thin, or Google-only

Review signals are consumed by AI systems through the platforms that publish them. A firm with 240 Google reviews and nothing anywhere else looks strong to a human scanning the map pack and looks incomplete to a system pulling from Yelp, Avvo, Justia, Facebook, and the BBB.

Recency matters as well. A 4.9 rating built entirely from reviews left in 2021 reads differently than a 4.7 with fresh reviews arriving every month.

7. Your firm exists nowhere except your own website

Ahrefs studied which brand-level signals correlate with AI visibility and found that branded web mentions correlated at roughly 0.66 to 0.71, while YouTube mentions showed the strongest single correlation at approximately 0.737.

The same study found something that should change how you budget. For ChatGPT specifically, correlation with classic authority metrics was weak, and the number of pages on a site had almost no relationship with AI visibility at all (approximately 0.194).

Publishing your 300th thin blog post is not going to fix this. Getting your managing partner quoted in the local business journal, cited in a state bar publication, interviewed on a regional podcast, and consistently named across YouTube and legal media might.

8. Legal is simply a harder category

Be realistic about the difficulty curve. Semrush and Growth Memo tracked 1,094 US categories in ChatGPT from January through June 2026 and found that legal and healthcare set a higher bar than other verticals. Even full prompt coverage and repeated citations did not reliably translate into brand-mention share.

Legal and medical queries are exactly where AI systems are most cautious. Your competitors face the same wall. The firms that get through it are the ones with real, verifiable, third-party-corroborated credibility, which is a slower build and a much more durable moat.

The real ChatGPT ranking factors for lawyers

Nobody outside OpenAI knows the weighting, and any agency that hands you a numbered list of "the 14 official ChatGPT ranking factors" is inventing it. What we do have is observable crawler behavior, published citation studies, and official platform documentation. Group them like this.

Layer one: retrievability. Can it reach you at all? OAI-SearchBot allowed in robots.txt and unblocked at the firewall. Pages indexed in Bing. Clean sitemap. Fast, server-rendered pages that do not hide core content behind JavaScript. No critical facts trapped in images or PDFs.

Layer two: entity clarity. Does the web agree on who you are? Identical firm name, address, and phone number everywhere. Correct practice area categories. LocalBusiness or LegalService schema markup with real values. Attorney bio pages with credentials, bar admissions, and jurisdictions.

Layer three: third-party corroboration. This is the heaviest layer for legal. Complete and current profiles on Google Business Profile, Yelp, Avvo, Justia, FindLaw, Lawyers.com, Facebook, and the BBB. Reviews across more than one platform, arriving steadily. Earned mentions in press, legal media, local news, and video.

Layer four: extractable substance. Pages that answer specific questions specifically. Clear headings. Direct opening sentences that answer the question before elaborating. Case results, verdicts, and settlements with context. Jurisdiction-specific detail. Named authors with credentials. Visible publication and update dates.

Layer five: freshness. Content that gets reviewed and updated, pushed through IndexNow so the current version is what gets read.
That is the honest picture. It is unglamorous, and it looks a great deal like doing excellent local SEO and reputation work with an emphasis shifted toward third-party platforms.

How to measure ChatGPT visibility without fooling yourself?

Prompt testing at volume. Build a list of 20 to 40 prompts a real client would use, in your practice areas and your city. Run each one repeatedly on a set schedule. Record how often your firm is named and which sources ChatGPT cited. Report your appearance rate as a percentage across many runs.

Referral traffic. OpenAI appends utm_source=chatgpt.com to referral URLs, so you can segment ChatGPT traffic in Google Analytics. Expect volumes to look small next to Google. That is normal and does not mean the channel is unimportant, because a meaningful share of AI-influenced prospects never click at all. They read the answer, then call you or search your firm name directly.

Bing Webmaster Tools AI Performance. Microsoft launched an AI Performance report in February 2026 showing which of your URLs get cited in AI-generated answers and the grounding queries behind them. To be precise, Microsoft describes its coverage as Copilot, AI-generated summaries in Bing, and select partner integrations, so treat it as a useful proxy rather than a direct ChatGPT report.

Your intake sheet. Add "ChatGPT or another AI assistant" to your how-did-you-hear-about-us options and actually train intake staff to ask. Most firms are already getting AI-influenced calls and logging them as "Google" or "referral."

Two things specific to running a law firm

Ethics rules still apply, and AI does not soften them. ABA Model Rule 7.1 prohibits false or misleading communications about a lawyer or the lawyer's services, and states diverge significantly from there. Several states require a prior results disclaimer next to any verdict or settlement figure, restrict testimonials, or mandate an "Attorney Advertising" label. Publishing more case results and testimonials is genuinely good for AI visibility, and it is also the fastest way to create a compliance problem. Have your compliance counsel review the pages before they go live, not after.

Paid placement in ChatGPT now exists, and it is separate from all of the above. OpenAI began testing ads in ChatGPT in the US in early 2026 and has since expanded to additional markets. Ads appear as clearly labeled sponsored units shown only to users on the Free and Go plans, and they sit apart from the organic answer. Buying ads does not make ChatGPT recommend you in its answer, and it is worth understanding that distinction before a vendor blurs it for you.

Frequently asked questions

Why does ChatGPT recommend my competitor and not me?

Usually because your competitor is easier to retrieve and easier to corroborate. They are indexed where the retrieval layer looks, present on the platforms it cites for legal (Google Business Profile, Yelp, Justia, Avvo), and their pages state specific, checkable facts. It is rarely because they are a better law firm.

How long does it take to become visible in ChatGPT?

Crawler and indexing fixes can show up in days. Directory and review work typically takes one to three months to settle. The third-party mention and authority layer is a six-to-twelve-month build. Anyone promising results in two weeks is describing the easy part only.

Do I need to abandon SEO and focus on AI?

No, and doing so would be an expensive mistake. Google is still used by 71.9% of US consumers researching attorneys, and strong organic rankings correlate with being named in AI answers. Your SEO investment now pays in two places instead of one.

Does having a blog help my ChatGPT visibility?

Good, specific, jurisdiction-aware content helps. Volume alone does not. The correlation between page count and AI visibility is close to nil.

Can I pay OpenAI to be recommended?

No. You can buy clearly labeled ad placements in ChatGPT, which are separate from the answer itself. The recommendation inside the answer is not for sale.

Is Bing really that important for ChatGPT?

Important enough to fix, yes. OpenAI has acknowledged using a third-party search provider alongside its own crawler; Bing has been the most widely documented one, and being indexed there costs you nothing but an afternoon. Treat it as a floor check rather than a ranking strategy.

The bottom line

The firms winning ChatGPT visibility right now are not the ones with the cleverest AI tricks. They are the ones that are easy to find, consistent everywhere, backed by real reviews on more than one platform, mentioned by people other than themselves, and specific enough on their own pages to be worth quoting.

That work was already worth doing. What changed is that there is now a second, fast-growing channel where it pays off, and a large group of your competitors is still buying llms.txt files instead of doing it.

Start with your robots.txt file today. It takes ten minutes, and for a surprising number of firms, it is the entire problem.