"Best estate planning attorney in [city]?" In 2026, ChatGPT, Perplexity, and Google AI answer that question in under five seconds and name one. So does the closely related high-stakes query: "Who should I hire for a wrongful termination case in [state]?" If your firm is not the answer, a competitor's is — and most practice owners never realize AI is making this call. The law firm AEO playbook is meaningfully different from individual-attorney SEO. Here it is: Attorney and LegalService schema properly nested, Avvo and Martindale-Hubbell profiles claimed and complete for every lawyer, NAP consistency across attorney-vs-firm sub-listings, and practice-area-specific FAQ content AI can cite directly.

Why AI Recommendations Are Reshaping How Clients Choose Legal Counsel

The structural shift that already hit retail, restaurants, and real estate is now propagating to legal. When a prospective client asks Perplexity "best personal injury attorney in Nashville" or asks ChatGPT "who should I contact about a probate matter in [state]," they are not browsing a directory — they are getting a synthesized recommendation. AI engines answer with the attorney name, plus a one-line citation: "Sarah Mitchell, Mitchell & Associates — 14 years active bar membership in Tennessee, 47 Avvo reviews, specialty tags in estate planning and probate."

The reason the law firm tier was late to AEO is structural. Individual attorneys see Avvo profile views directly; firm-level signals (NAP consistency across attorney sub-listings, LegalService schema nesting, state-bar directory coverage, firm-level practice-area content) were largely invisible. That has changed. AirOps recency data shows that legal-intent queries have a 30–60 day recency window — tighter than the 90-day window that works for many SMB verticals — because clients looking for counsel want the practice that is taking cases now, not the practice that was hot a year ago. A firm with stale aggregator signals gets skipped at the firm-tier query regardless of how well individual attorneys rank on their own profiles.

How AI Engines Discover and Recommend Law Firms

AI engines for legal services pull from a layered citation graph that is meaningfully different from other SMB verticals. The legal-vertical layers are:

Legal Directories — Avvo, Martindale-Hubbell, Justia, FindLaw, Lawyers.com, Super Lawyers

Avvo and Martindale-Hubbell are the highest-authority aggregator sources for almost any query that touches attorney data. AI engines read attorney-level pages on these platforms (ratings, specialty tags, review counts, jurisdictional admissions, disciplinary record) and weight them heavily for tier-1 attorney queries. Attorney data on Avvo and Martindale significantly outweighs data on FindLaw and Lawyers.com for the same query — so even if every attorney on your team has a complete FindLaw profile, the Avvo and Martindale layers are what AI cites for "best [specialty] attorney in [city]." Justia is weighted for legal-information-adjacent queries (the platform carries attorney profiles plus a public legal-info library that AI cites directly). Super Lawyers and Best Lawyers published lists carry disproportionate weight at the tier-2 attorney query — getting on those lists and keeping your profile current there is one of the highest-leverage third-party signals in this vertical.

Google Business Profile With Attorney vs. Firm NAP

The NAP (Name, Address, Phone) drift problem is particularly acute in legal because most firms run dual presences: the firm-level Google Business Profile for the office, plus individual attorney-level listings some senior attorneys maintain separately. When the firm NAP and the attorney NAP differ in suite number, phone line, or attribution formatting, AI engines de-duplicate inconsistently and recommend the lawyer individually while skipping the firm. This is the same failure mode that costs brokerages their tier-1 brokerage queries — for law firms, it costs the firm-tier query. The fix is purely operational: pick one canonical NAP per office and per attorney surface, then enforce byte-identical consistency across every directory listing.

State Bar Association and State-Court Directories

State bar directories are the only sources that verify current active bar membership and disciplinary status, and AI engines treat them as primary-source verification. The state-bar citation is treated by ChatGPT and Perplexity as the equivalent of a Wikipedia citation for the attorney — the canonical reference point. Less weight but still cited: federal court CM/ECF attorney admissions, state-court registry listings, the U.S. Patent and Trademark Office patent-attorney roster (for IP specialists), and the equivalent bar regulator in every other state. The fix is to make sure every practicing attorney has a complete, current listing on their state-bar directory.

Firm Website With Attorney + LegalService Schema

Your firm website is the canonical source for the LegalService entity that wraps your entire practice. Proper markup — one LegalService per office, nested Attorney members, plus per-practice-area FAQPage schema — gives AI engines a single connected graph to walk at machine-reading speed. Add practice-area FAQ content answering the questions clients actually ask ("how much does a probate attorney cost in [state]?", "what is the statute of limitations for wrongful termination in [state]?", "do I need an attorney for a real estate closing in [state]?") and the recommendation likelihood compounds sharply because the AI can cite your firm as the source for both the recommendation and the substantive answer.

Third-Party Citations — Local Media, Award Lists, Case-Result Publications

Firm-level mentions in local press — "Mitchell & Associates secured a $4.2M verdict in [county] last month" — carry more AEO weight at the firm tier than solo mentions do at the individual tier. Same story for "best of [city]" listicles in regional magazines, bar journal publications, and the case-result publications on your firm's website that AI engines can verify through court records. These are third-party credibility signals AI engines cannot be self-manufactured, and they cite them disproportionately for tier-1 firm-tier queries.

5 Actionable Steps for Law Firm AEO

Step 1: Claim and complete Avvo and Martindale-Hubbell profiles for every attorney

Every attorney at the firm needs a current, complete Avvo profile and a current Martindale-Hubbell profile. These two directories are the highest-weight aggregator sources for the legal vertical. Most firms have profiles but leave them at 40–60% completion, missing the fields AI engines weight most heavily.

Each attorney profile must include:

Perplexity and ChatGPT both pull from Avvo and Martindale-Hubbell when making attorney recommendations. A profile that is 80% complete and updated in the past 60 days beats one that is 100% complete but has not been touched in two years — the recency shift on these directories has been sharp through 2025 and 2026.

Step 2: Audit and unify attorney-vs-firm NAP across every directory

This is the highest-leverage operational move for the law firm tier. Programmatic NAP audit using BrightLocal, Whitespark, or Yext: the output is one canonical NAP per office and per attorney listing, with a list of every directory surface where the NAP currently drifts. Drift here typically appears on the suite number, phone line, or attorney-vs-firm attribution string (e.g., "Mitchell & Associates" vs. "Mitchell and Associates" vs. "Mitchell & Associates, Attorneys at Law" — three different listings, one firm). That drift is what causes AI engines to skip the firm and recommend individual attorneys only.

Include in the audit every directory surface: Avvo attorney page, Martindale-Hubbell attorney page, Justia attorney profile, FindLaw attorney profile, Lawyers.com attorney profile, Google Business Profile for the firm office, individual attorney GBPs where they exist, Apple Maps, Bing Places, Yelp, Facebook business page, LinkedIn company page, and the long tail of aggregator and legal-directories. The audit typically surfaces drift on 25–40% of surfaces, often on the attorney-vs-firm attribution strings the firm operations teams never touch.

Step 3: Add Attorney + LegalService JSON-LD with member graph on the firm website

Schema nesting is the pattern that wins the firm-tier query. One LegalService entity per firm, with a member array pointing at every Attorney entity. Then one Attorney schema per lawyer's page, with worksFor pointing back to the LegalService. The two types live side by side on your site, and the connections between them are what AI engines walk when answering a firm-tier query.

A proper LegalService schema includes:

An Attorney schema for each lawyer includes: @type: Attorney; name; jobTitle; worksFor (pointing back to the LegalService entity); address; telephone; alumniOf; knowsAbout (lawyer-specific specialty tags); sameAs (Avvo, Martindale, Justia, Lawyers.com profile URLs). Most legal websites have one of the two schema types but not both in a connected graph — the connection is what does the work. Test at Google's Rich Results Test after implementation; errors mean that field is being ignored by AI engines, not just that Google alone is dropping it.

Step 4: Build practice-area FAQPage content with state-specific answers

Practice-area queries are where AI directly cites firm websites in the legal vertical. Queries like "how much does probate cost in [state]?", "what is the statute of limitations for wrongful termination in [state]?", "do I need a lawyer for a real estate closing in [state]?", "what is the difference between Chapter 7 and Chapter 13 bankruptcy in [state]?" — these are queries where ChatGPT, Perplexity, and Google AI Overviews pull their substantive answers from firm-website FAQ content, then cite the firm as the source.

Each practice-area FAQ page should answer:

The content density target is one FAQ page per practice area the firm actively serves, plus a state-specific version of each frequently asked question where the answer materially differs by jurisdiction. Specific numbers — actual fee ranges, actual statute-of-limitations deadlines, actual court timelines — get cited. Vague disclaimers ("every case is different, consult an attorney") do not. AI engines cite pages where the substantive answer is on the page itself, not pages that defer to a consultation.

Step 5: Build review velocity on GBP and Avvo with recency framing

Review velocity is a separate signal from review count. AI engines read the recency-weighted review volume on Google Business Profile and Avvo as a distinct credibility signal at the firm and attorney tier. Your target is not 4.9 stars across 800 reviews aggregated over five years — it is a steady inflow of 3–5 reviews per attorney per month on Google plus Avvo combined, with most reviews dated within the past 90 days.

Aim for at least four new reviews per month per attorney across GBP and Avvo combined. The soliciting mechanism is straightforward: the request goes out at case resolution with a direct link to the GBP or Avvo review form, not a generic "leave us a review" link to a directory of the firm's choosing. Track the review count by source by attorney in the firm CRM or case management system; attribution per attorney matters because AI engines read each attorney's review stream independently.

Find out how visible your law firm is to AI right now.
Free AEO scanner — checks ChatGPT, Perplexity, and Google AI visibility in 60 seconds.

Run Your Free AEO Scan →

What AI Recommends vs. What It Misses for Law Firms

What AI recommends: Firms with complete Avvo and Martindale-Hubbell profiles for every practicing attorney, state-bar directory listings verified as active good-standing for every lawyer, LegalService schema wrapping every Attorney in a connected member graph, practice-area FAQ content with state-specific answers AI can cite directly, and firm-level review velocity on GBP plus Avvo of at least 30+ reviews aggregated across the firm in the past 12 months. Specialty firm positioning — estate planning, personal injury, employment law, family law, immigration, IP — gets disproportionate AI share for the specialty query set because the specialization signal is easier for AI to confirm across multiple data sources, and clients searching for a specialty typically want firm-tier answers, not solo attorneys.

What AI misses: Firms with strong individual-attorney authority but no firm-level signals (the parent practice is invisible even when every lawyer on the team ranks individually on Avvo); firms running attorney-vs-firm listings with NAP drift across directory surfaces; firms with state-bar directory listings that are out of date or missing practice-area registration; firms whose practice-area FAQ content is generic boilerplate without state-specific answers; multi-office firms with cross-office NAP drift (the same firm name appears as "[Firm Name] — [City A]" in one jurisdiction and "[Firm Name] LLC" in another, and AI engines do not connect them). The gap is wide. A firm that consistently works its directory completeness, NAP consistency, schema nesting, and practice-area FAQ content wins the tier-1 firm query for its specialty within 60 to 90 days, even against better-known competitors with no firm-tier AEO investment.

Go Deeper