What changes when the answer appears before the click — and how firms can remain the cited source.
The answer moved above the results
For twenty years, legal search worked one way: a prospective client typed a question, scanned ten blue links, and clicked two or three. Google's AI Overviews — and answer engines like ChatGPT, Perplexity and Gemini — have inserted a new layer above those links. The engine now reads the pages, synthesizes an answer, and cites a small number of sources. For many informational queries, the client gets what they need without ever scrolling to the traditional results.
This does not kill search traffic for law firms; it concentrates it. The firms cited inside the AI answer capture a disproportionate share of the clicks that remain, and — just as important — they shape the client's shortlist before any comparison shopping begins. Being source number three in an AI Overview is worth more than being link number three below it.
What answer engines actually cite
Across the legal queries we track, cited sources share consistent traits. They answer a specific question directly in the first paragraph, then expand. They use clear structural markup — headings that mirror real questions, concise definitions, step lists and tables. They demonstrate first-hand expertise: named attorneys, bar admissions, jurisdiction-specific detail, and publication dates that are kept honest.
Just as decisive is entity clarity. Engines cite sources they can confidently attribute. A firm whose name, attorneys, practice areas and locations are consistent across its website, bar directories, Google Business Profile and major legal directories is far easier for a model to trust than one with fragmented, contradictory listings.
What firms should do now
Start with the questions your intake team hears every week. Each one deserves a page or section that answers it plainly in the first two sentences, written for a worried person rather than a law review. Add the structured data that disambiguates your firm — Organization, Attorney, LegalService and FAQ markup — and make sure every attorney bio carries credentials a machine can parse.
Then consolidate authority. Fewer, deeper pages outperform dozens of thin practice-area variants. A single authoritative guide to car accident claims in your state, kept current, earns citations that fifty near-duplicate location pages never will.
How we measure AI visibility
Traditional rank trackers cannot see AI answers, so we monitor them directly: a fixed panel of high-intent queries per practice area, checked across the major answer engines, recording whether the firm is cited, which page is cited, and how the answer frames the firm. That panel becomes a baseline we report against monthly, alongside the classic metrics of rankings, traffic and signed cases.
The firms that treat AI visibility as a measured channel — rather than a curiosity — are the ones compounding an advantage while their competitors wait for the old playbook to start working again.
The queries that matter most
Not every search is worth chasing into the AI layer. The queries that decide retainers are the high-intent, high-anxiety ones: 'what happens after a DUI arrest in Florida,' 'how long do I have to file a malpractice claim,' 'do I need a lawyer for a custody modification.' These are exactly the questions answer engines were built to resolve — and exactly the pages most firm websites either lack or bury under marketing language.
Build the inventory from your own intake logs. Every question a receptionist or intake specialist answers twice deserves a page. Group them by practice area and jurisdiction, because 'statute of limitations' means different things in different states, and engines reward the page that is specific about which one it is answering.
Structured data that earns trust
Schema markup is how you tell a machine, in its own vocabulary, what your pages mean. For a law firm the essential types are Organization and LegalService on the homepage, Attorney on every bio, FAQPage on genuine question-and-answer content, and BreadcrumbList everywhere. None of this is decorative: it is the evidence layer engines use to decide whether they can safely cite you.
The discipline is honesty. Mark up only what is visibly on the page, keep it in sync when attorneys join or leave, and validate it after every site change. Structured data that contradicts the visible page is worse than none at all — it teaches engines that your markup cannot be trusted.
Keeping content citable over time
Answer engines prefer current sources for time-sensitive questions, and legal questions are almost always time-sensitive. A guide to filing deadlines that still describes a repealed rule does not just rank worse — it teaches engines that your site requires verification they would rather skip. Every substantive page needs a visible 'last reviewed' date and an owner responsible for keeping it true.
Set a review cadence by practice area: quarterly for anything involving deadlines, fees or procedures, annually for evergreen explainers. When the law changes, update the existing page rather than publishing a new one — the accumulated authority and citations stay with the URL.
From rankings to citations: a practical GEO playbook
Generative engine optimization is not a replacement for SEO — it is the same discipline extended to a new surface. The technical foundations still decide everything: crawlable pages, fast rendering, honest structured data and content that answers real questions. What changes is the unit of success. Classic SEO chases positions; GEO chases citations — the moments an engine names your firm, quotes your page, or recommends your attorneys inside a synthesized answer.
In practice that means three habits. First, write answer-first content: every important question your intake team hears gets a page that resolves it in the opening sentences, then supports it with jurisdiction-specific detail. Second, keep your entity graph clean so models can attribute what they read — consistent names, bar admissions, locations and profiles everywhere your firm appears. Third, monitor the engines themselves, because neither rankings nor traffic reports will tell you when ChatGPT starts recommending a competitor by name.
The window of advantage is open now precisely because most firms are waiting. Answer engines build citation habits around sources they already trust, and dislodging an entrenched source is far harder than becoming one early.
Law firm growth intelligence, monthly.
One concise email a month: what is changing in legal search, AI visibility and intake — and what to do about it. No spam, unsubscribe anytime.
We use your email only to send the briefing. See our Privacy Notice.
