GEO for Law Firms: Why Seven Directories Own Your AI Citations

The referral business that stopped being a referral business

Legal has run on referrals and reputation for a century. A client asks a friend, the friend names a firm, the firm's name carries weight because someone vouched for it. That system is being routed through a new middleman, and most firms haven't noticed.

79% of legal professionals now use AI tools internally, 87% at large firms, 71% even at solo practices, according to Clio's 2025 Legal Trends Report as cited in a 2026 report from 5W Public Relations and the Haute Lawyer Network. Lawyers have adopted AI for drafting, research, and internal workflow faster than almost any other profession. What they haven't done, in most cases, is check whether AI can find and recommend them to a prospective client.

That gap matters because the client side has already moved. 23.6% of legal search queries now trigger Google AI Overviews, and for question-style queries, the kind a worried person types at 11pm after an accident or an arrest, that figure rises to 57.9%, per an Ahrefs analysis of 146 million SERPs cited in the same report. A meaningful share of the profession's highest-intent traffic no longer lands on ten blue links. It lands on an AI-generated answer, and that answer cites a source.

The uncomfortable finding, across two independently run studies, is that the source is almost never the law firm itself.

Who actually owns the legal AI citation layer

The 5W/Haute Lawyer report names the problem plainly: when ChatGPT, Claude, Perplexity, or Google AI Mode recommend a lawyer, the answer traces back to one of seven directories, Chambers, Legal 500, Super Lawyers, Best Lawyers, Martindale, Avvo, and Justia. Individual firms, including some of the most prestigious in the country, show up inside those directory profiles rather than as sources in their own right. The report's own test case is blunt: a query for "Cravath, Swaine & Moore M&A expertise insights," a question that should surface Cravath's own thought leadership, returned two Chambers profiles, two Cravath practice pages, a Legal 500 profile, and a Chambers Global profile in the top six results. The directories out-cited the firm on the firm's own specialty.

Ronn Torossian, 5W's founder and chairman, put the underlying dynamic this way: "The seven directories that currently own legal AI citations are extracting fees from a discovery layer the industry built for them by default."

InterCore Research's independent study, published July 2026, quantifies the same pattern from a different angle. Its methodology is unusually rigorous for this kind of report: 540 high-intent legal queries across six practice areas and 30 US metros, three phrasings each, each combination run three times, for 1,620 total AI answers and 18,900 individual source citations, all tested against Perplexity Sonar and measured on a single date (July 20, 2026). The full dataset is published, which is worth noting on its own, most AI-visibility studies aren't reproducible.

The headline numbers: a legal directory was the first source cited in 77.8% of answers (1,254 of 1,612), and directories account for at least 51.8% of all 18,900 citations, a figure InterCore reports as a floor rather than a point estimate because 36% of citation volume sits on domains not yet classified, some of which are themselves further directories. Justia alone was cited in roughly two-thirds of answers. InterCore's own name for this pattern is "the directory wall."

Firm-level citation, by contrast, is fragmented almost past the point of usefulness. 254 different law firm websites split what's left after directories take their share, and the single most-cited individual firm appeared just 36 times across all 18,900 citations, under 0.2%. No firm has built enough independent citation authority to be a consistent, repeatable AI answer. Directories win by default because nobody else has assembled a comparable footprint.

The sharper problem: named without being linked

There's a distinction inside this data that most firms will miss if they only skim the headline stats, and it's the single most actionable number in either study.

This is a different failure mode from simply not being visible. A firm with no AI visibility at all has a content and authority problem, full stop, that's the harder, slower fix. A firm that's named but not linked has something narrower and more fixable: it has enough reputation to be recalled, but nothing citable enough to be pointed to. Closing that specific gap, giving AI a page to link to instead of just a name to recall, is lower-hanging fruit than most firms realize.

Part 1: Directory presence is table stakes, not strategy

Given that seven directories capture the large majority of first-cited legal sources, a complete, accurate, well-differentiated profile on Chambers, Legal 500, Super Lawyers, Best Lawyers, Martindale, Avvo, and Justia is no longer optional groundwork. It's closer to the price of entry.

That means, per firm and per named attorney: consistent name, practice area, and location data across all seven (AI models penalize inconsistency the same way local search does), complete bio fields rather than the bare minimum a directory requires, verified bar admissions and years of practice, and specific practice-area tagging rather than generic "litigation" or "corporate" labels that blur into every other listing. Most firms treat directory profiles as a one-time submission. Given how heavily AI leans on these seven sources, they deserve the same maintenance discipline as a firm's own website.

Directory presence gets you into the citation pool. It doesn't differentiate you inside it, every competitor on your own list is on the same seven directories. The differentiated opportunity is everything below.

Part 2: Practice-area content that AI can actually cite

Directories win the "first cited" spot because they're structured, deep on every practice area, and easy for a retrieval system to parse. Most law firm websites are the opposite: a handful of thin service pages, no real depth per practice area, nothing that answers the specific question a prospective client typed into ChatGPT.

Build a distinct, substantial page per practice area, not one page listing every service the firm offers. Each page should answer the actual questions people ask before hiring: what does this process look like, what does it typically cost or how is it billed, what's a realistic timeline, what mistakes do people make before contacting a lawyer. Answer the core question in the first two sentences, the same way you'd want an associate to open a client memo, not bury it three paragraphs into a "why choose us" section.

Case results pages, where your jurisdiction's bar advertising rules permit them, are some of the most citable content a firm can publish, specific, factual, and exactly the kind of concrete detail retrieval systems favor over marketing copy. Bar rules on describing outcomes vary significantly by state, some require disclaimers, some restrict specific dollar figures or "similar results" language, so check your state bar's advertising rules before publishing (this isn't legal advice, and it shouldn't need to be said to a room full of lawyers, but the AI-visibility incentive to publish results makes it worth restating: compliance comes first).

FAQ sections built around the actual question-style queries your prospects ask, the kind driving that 57.9% AI Overview trigger rate, are a direct lever. If you serve any local market alongside your practice-area focus, the content and entity-clarity fundamentals in GEO for local businesses apply directly, especially for firms with a strong local-search component to their client acquisition.

Part 3: Attorney bio pages and credential schema

An attorney bio page is one of the few asset types a directory profile can never fully replicate, because only the firm can publish the full, first-person version of a lawyer's record. Treat it that way.

Structure bio pages with Person schema carrying the fields that actually matter for legal trust signals:

{
  "@context": "https://schema.org",
  "@type": "Person",
  "name": "Attorney Name",
  "jobTitle": "Partner",
  "worksFor": {
    "@type": "Attorney",
    "name": "Firm Name",
    "url": "https://firmdomain.com"
  },
  "alumniOf": {
    "@type": "CollegeOrUniversity",
    "name": "Law School Name"
  },
  "hasCredential": {
    "@type": "EducationalOccupationalCredential",
    "credentialCategory": "Bar Admission",
    "recognizedBy": {
      "@type": "Organization",
      "name": "State Bar Name"
    }
  },
  "knowsAbout": ["Practice Area One", "Practice Area Two"]
}

Beyond schema, the content on the page itself should carry the specifics AI systems extract as trust signals: bar admissions with the actual jurisdictions and dates, notable case results (subject to the same state-by-state advertising rules noted above), speaking engagements, published articles, and years handling the specific practice area a prospective client is researching. "Experienced attorney" is not extractable. "Admitted to the New York and New Jersey bars in 2014, has handled over 200 employment discrimination cases" is.

This is also where the seven-directory dependency gets diluted over time. A directory profile is a summary written to that directory's template. An attorney bio page, done properly, is the firm's own primary source, and it's the asset most likely to close the 34.2% named-but-not-linked gap for a given attorney.

Part 4: Entity clarity for multi-office and local firms

Many firms, especially regional and multi-office practices, are simultaneously a professional services entity and a local business, and AI systems need both identities to resolve cleanly. A firm with three offices needs three distinct, consistent local entities, not one generic "contact us" page that mixes every location's information together.

Use LegalService or Attorney schema per location, with consistent NAP (name, address, phone) data matched exactly across the website, Google Business Profile, and all seven directories, the same consistency principle that governs local SEO generally, and covered in more depth in GEO for local businesses. For firms competing on referral relationships with corporate clients rather than consumer intake, the entity-authority and content-architecture approach in GEO for B2B is a closer structural match than the consumer-intent playbook above.

Measuring legal AI visibility

90-day roadmap

Month 1: Directory and entity foundation

• Audit and complete profiles on all seven directories (Chambers, Legal 500, Super Lawyers, Best Lawyers, Martindale, Avvo, Justia) for every named attorney
• Fix NAP consistency across website, Google Business Profile, and directories, per office if multi-location
• Implement Person and Attorney/LegalService schema across bio and location pages
• Run manual query sampling across ChatGPT, Perplexity, and Google AI Mode for your top 10 practice-area terms to establish a baseline

Month 2: Content that closes the citation gap

• Build or rebuild dedicated practice-area pages with question-first structure
• Publish attorney bio pages with full credential detail and Person schema
• Confirm bar advertising rules per state, then publish compliant case results and outcome pages
• Add FAQ sections targeting the specific question-style queries clients ask before contacting a firm

Month 3: Close the named-but-not-linked gap

• Re-run the query sampling from Month 1 and identify which named-but-uncited attorneys and practice areas still lack a linkable page
• Prioritize content for those specific gaps first, this is the highest-leverage fix in the whole roadmap
• Pursue coverage in legal press and industry publications as a secondary authority signal
• Set a recurring quarterly cadence for directory and schema maintenance, this is not a one-time project

Frequently asked questions

Which directories matter most for legal AI citations?

Two independent 2026 studies point to the same seven: Chambers, Legal 500, Super Lawyers, Best Lawyers, Martindale, Avvo, and Justia. A 5W Public Relations and Haute Lawyer Network report found these seven own the large majority of first-cited sources when AI models recommend a lawyer, and InterCore Research separately found Justia alone cited in roughly two-thirds of AI answers to legal-hiring queries. Complete, accurate, consistent profiles on all seven are close to a prerequisite for legal AI visibility.

What does "named but not linked" mean, and why does it matter?

It's InterCore Research's finding that 34.2% of law firms AI names in an answer have no page of their own among the sources cited for that answer. The AI recalls the firm's existence, likely from directory listings or training data, but has nothing of the firm's own content to point to. It's a distinct problem from having no visibility at all: these firms already have enough reputation to be recalled by name. What they lack is a citable page, an attorney bio, a practice-area explainer, a case results page, that AI can actually link to.

Can law firms publish case results for AI citation purposes?

Only where state bar advertising rules permit it, and those rules vary significantly by jurisdiction, some require specific disclaimers, some restrict describing outcomes as "typical" or comparing them to other cases. Check your state bar's advertising rules before publishing anything result-specific. This isn't legal advice. Where permitted, case results pages tend to be highly citable because they're specific and factual, exactly what AI retrieval systems favor over general marketing language.

How is legal AI visibility different from a local SEO problem?

They overlap but aren't identical. Local SEO governs how a firm's office locations resolve in Google Maps and local pack results. Legal AI visibility, per InterCore's directory-wall finding, is more about which source AI cites first when someone asks a hiring question, and that's dominated by national directories, not local listings. Firms with a strong local-search component to client acquisition should treat the two as complementary, not the same project, our GEO for local businesses guide covers the local-specific fundamentals.

How often should a firm re-check its AI citation status?

Quarterly at minimum, given how fast AI retrieval indexes update and how competitive the seven-directory citation layer is. Re-run the same set of practice-area and attorney-name queries each quarter across ChatGPT, Perplexity, and Google AI Mode, and track whether previously named-but-unlinked attorneys have gained a citable page.

Sources: 5W Public Relations and Haute Lawyer Network, "79% of Lawyers Use AI Internally", April 2026, InterCore Research, "The State of AI Search Visibility for Law Firms (2026)", July 2026