41% of legal searches now involve AI in some form. How law firms get cited by ChatGPT, Perplexity, and Google AI Overviews. AEO and GEO strategies that work today.
By the numbers
Something shifted in legal search over the past 18 months. Not a tremor. A tectonic plate movement. 41% of legal searches now touch an AI platform in some form. ChatGPT. Perplexity. Google AI Overviews. Microsoft Copilot. When a potential client asks "best personal injury lawyer in Chicago," the answer used to be ten blue links. Now it's often a direct AI-generated recommendation — one firm, maybe two, with a brief explanation of why.
Let's get specific about what's happening. When someone types a legal question into ChatGPT ("Do I need a lawyer after a car accident?" or "How does child custody work in Texas?"), the response isn't a list of links. It's a direct answer. And increasingly, that answer includes a recommendation. A firm name. A link. A reason why.
Perplexity does the same thing, but with real-time web search baked in. Google AI Overviews (the AI-generated summaries that now appear above organic results for many queries) pull directly from indexed pages and synthesize answers on the spot. Copilot draws from Bing's index. Each platform works differently under the hood, but the user behavior is identical: ask a question, get a direct answer, skip the search results page entirely.
Here's the part most law firms miss: there's no "position 7" in a ChatGPT response. Traditional SEO has a spectrum: you can be on page one, page two, position three, position fifteen. Visibility degrades gradually. AI search is binary. You're either cited or you're not. You're either the recommended firm or you don't exist in that interaction. That's a fundamentally different competitive dynamic, and it rewards a different kind of optimization.
The 41% number will surprise some people. It shouldn't. Google reported that AI Overviews alone appear for over 30% of search queries. Layer on ChatGPT's 200 million weekly active users, Perplexity's rapid growth, and Copilot's integration into every Windows device, and the picture is clear. Our analysis of AI Overviews breaks down exactly how this impacts law firm visibility in organic search. For a broader look at how this is reshaping client acquisition, read our piece on how AI is changing the way people find lawyers.
Traditional SEO targets Google's index. You optimize pages so Google ranks them. AEO (Answer Engine Optimization) targets the AI models that generate answers. Different target, different rules.
The core distinction: Google ranks pages on a list. AI models cite sources in a generated response. There's no list. There's no ranking. The AI either pulls from your content and mentions your firm, or it pulls from someone else's. That's the entire game.
| Approach | Target | How it works | Outcome |
|---|---|---|---|
| SEO | Google's ranked results | Optimize pages for keywords, links, and technical health | Higher position on results page |
| AEO | AI-generated answers | Structure content so AI models select it as the answer | Your firm is cited in the direct answer |
| GEO | AI source selection | Build authority signals and structured data AI trusts | Your content becomes the source AI pulls from |
AEO isn't replacing SEO. Think of it as a new layer on top of your existing strategy. The foundation (quality content, technical soundness, domain authority) still matters. But AEO adds a set of optimization targets that traditional SEO doesn't address:
Factual density. AI models prefer content packed with specific, verifiable facts over vague marketing language. "Our attorneys are dedicated to fighting for your rights" gives an AI nothing to cite. "Founded in 2003, the firm has resolved 2,400+ personal injury cases with average settlements 34% above the county median" gives it everything. Numbers. Dates. Specifics. That's quotable material.
Third-person authority signals. AI models are trained to distrust self-promotional language and trust neutral, encyclopedic language. The same reason Wikipedia gets cited constantly and promotional landing pages don't. Your content needs sections written in third-person factual tone. Not your entire site, but strategic paragraphs that AI can extract without sounding like an advertisement.
Direct answer formatting. AI models scan for content that directly answers questions. FAQ structures, TL;DR paragraphs, definition blocks. These are the formats that get pulled into AI-generated responses. If your page buries the answer under 800 words of preamble, you lose to the competitor who puts the answer in the first paragraph.
GEO: generative engine optimization
If AEO is about being the answer, GEO is about being the source. Generative Engine Optimization focuses on how AI models decide which websites to cite when generating responses.
Structured data is the bridge. Schema markup (FAQ schema, Attorney schema, LocalBusiness schema, LegalService schema) translates your content into machine-readable format. AI models can parse unstructured text, but they prefer structured data because it's unambiguous. A page with proper schema is significantly more likely to be cited than an identical page without it. Use our schema generator to create valid JSON-LD for your firm.
Topical authority matters more than individual page optimization. AI models assess whether a source is authoritative on a topic, not just whether a single page answers a question. A law firm with 40 pages of interconnected content about personal injury law (from case types to settlement timelines to specific injury categories) signals topical authority that a firm with one generic practice area page can't match. Google's own guidance on page experience signals confirms that technical health amplifies these authority signals.
Citation-worthy writing has a specific anatomy. Short, factual paragraphs (60-90 words). Neutral tone in key passages. Specific data points rather than generalized claims. Sourced statistics. Clear attribution. These are the textual patterns that AI models select for citation. Marketing copy gets ignored. Encyclopedia-style content gets quoted. Our GEO guide for ChatGPT and Perplexity goes deeper into the specific tactics that earn citations on each platform.
Not all AI platforms work the same way. Each one has different data sources, different citation preferences, and different optimization levers. Here's the platform-by-platform breakdown from two years of testing.
ChatGPT. OpenAI's model draws from two sources: its training data (a massive snapshot of the web, legal directories, news archives, and professional databases) and real-time browsing (when the user enables it or when the model decides it needs current information). ChatGPT favors authoritative, well-structured content from established domains. It tends to cite firms that have strong presences on legal directories like Avvo, Martindale-Hubbell, and Super Lawyers, because those platforms are heavily represented in its training data. Sites with proper LegalService schema give ChatGPT's browsing mode structured data to pull from directly.
Perplexity. This is the platform most law firms are sleeping on. Perplexity performs real-time web searches for every query, then synthesizes results with source citations. It favors pages with clear, direct answers, especially pages with FAQ structure, schema markup, and well-organized content. Because Perplexity searches the live web, your current SEO performance directly influences your Perplexity visibility. But Perplexity is pickier about which parts of a page it quotes. It strongly prefers concise, factual paragraphs over long-form marketing content.
Google AI Overviews. These AI-generated summaries appear above traditional organic results for an increasing percentage of queries. AI Overviews pull from Google's own index, which means your existing SEO work directly feeds into them. But the selection criteria are different from organic ranking. AI Overviews strongly favor content that meets Google's helpful content guidelines: direct definitions, numbered steps, comparison tables, and FAQ answers. Our AI Overviews analysis has the full data on which content formats get selected most often.
Microsoft Copilot. Built on the Bing index, Copilot is the AI assistant integrated into Windows, Edge, and Microsoft 365. It favors Bing Places listings (the equivalent of Google Business Profile), structured data, and content from domains with high authority in Bing's ranking system. If you've been ignoring Bing SEO (and most law firms have), you're missing Copilot entirely. The good news: Bing is less competitive than Google, so the barrier to visibility is lower. Claiming and optimizing your Bing Places listing is a 30-minute task that can produce outsized AI visibility returns.
Structured content wins. Factual density wins. Schema markup wins. Marketing fluff loses. Every time. The firms using structured, question-and-answer content on their own sites are building exactly the kind of material these platforms love to cite.
Here's the step-by-step action plan we use with our clients. We've refined this across 200+ campaigns, and the order matters. Each step builds on the previous one.
Step 1: Audit your current AI visibility. Before you optimize anything, you need a baseline. Search your firm name in ChatGPT, Perplexity, and Google (with AI Overviews). Then search your practice area + location queries: "best personal injury lawyer in [your city]," "divorce attorney near [your area]," "how to find a criminal defense lawyer in [your state]." Document what comes up. If you're not mentioned anywhere, that's your starting point.
Step 2: Build a Brand Facts page. Create a dedicated page on your site with neutral, factual information about your firm. Founding year, number of attorneys, practice areas served, geographic coverage, notable case results (with specifics), professional affiliations, awards, and community involvement. Write it in third person. Keep it factual. No marketing language. This single page will become the primary source AI models pull from when generating information about your firm.
Step 3: Add FAQ schema to every page. Every practice area page, location page, and major blog post should have 6-12 FAQs with proper schema markup. The answers should be specific, factual, and self-contained. Each answer should make sense without reading the surrounding page. This is the format all four major AI platforms prefer when selecting content to cite.
Step 4: Write TL;DR opening paragraphs on service pages. Add a 60-90 word factual summary at the top of each practice area page. Neutral tone, specific numbers, direct statements. This is the paragraph AI models are most likely to quote when a user asks about your practice area.
Step 5: Get listed on authoritative legal directories. ChatGPT's training data includes massive amounts of directory content. Avvo, Martindale-Hubbell, Super Lawyers, FindLaw, Justia, Lawyers.com. As the American Bar Association and other legal bodies increasingly reference digital presence, these aren't just referral sources anymore. They're training data sources. Complete, detailed profiles on these platforms increase the likelihood of AI citation.
Step 6: Build topical authority through content clusters. AI models assess source authority at the topic level, not the page level. A single blog post about "car accident settlements" doesn't establish authority. Twenty interconnected pieces covering accident types, settlement timelines, liability factors, insurance negotiations, and state-specific laws? That signals authority. Build content clusters around your core practice areas with internal linking that maps the relationships between topics.
What makes content "quotable" by AI?
We've analyzed over 3,000 AI-generated responses about law firms across ChatGPT, Perplexity, and Google AI Overviews. The content that gets cited shares a consistent DNA. And it looks nothing like typical law firm marketing copy.
TL;DR paragraphs. The single most effective format for AI citation is a 60-90 word paragraph that answers a question directly, in neutral factual tone, with specific data points. Example that gets cited: "Smith & Associates is a Chicago-based personal injury firm founded in 2003. The firm's 12 attorneys have handled over 2,400 cases, recovering more than $180 million in total settlements." Example that never gets cited: "At Smith & Associates, we fight tirelessly for the justice you deserve."
FAQ structure. Every practice area page should have 8-12 FAQs with specific, direct answers. Not "Contact us to learn more." Actual answers. "The average timeline for a personal injury settlement in Illinois is 12-18 months from filing. Cases involving disputed liability or severe injuries may take 24+ months." That's the kind of answer AI quotes verbatim.
Brand Facts pages. A Wikipedia-style page on your own website that presents your firm's key information in neutral, third-person, factual format. It feels unusual to write about your own firm in third person, but AI models strongly prefer this format when generating firm recommendations.
One point to keep clear: this doesn't mean your entire website should read like a Wikipedia article. Your practice area pages still need compelling copy that converts human visitors. But you need both: persuasive content for the humans who land on your site, and factual structured content for the AI models that decide whether to send them there. Check out our resources library for practice-area-specific optimization guides.
Nothing dramatic. Not at first. That's the dangerous part.
You'll still get organic leads from traditional search. Your Google Business Profile will still generate calls. Referrals will still come in. Everything will feel normal, while a growing percentage of potential clients never see your firm's name because they got their answer from an AI instead of scrolling through search results.
The erosion is gradual. Then it isn't.
41% is today's number. In 18 months, it'll be 60%+. That's not speculation; it's the trajectory of every AI platform's user growth combined with Google's aggressive expansion of AI Overviews. When 60% of legal searches are influenced by AI, the firms not showing up in AI responses aren't losing a niche channel. They're losing the majority of their potential visibility.
We saw this exact dynamic play out with SEO adoption in 2008-2012. The firms that invested early built compound advantages that are still generating returns today. The firms that waited until 2015 or 2018 had to spend 3-5x more to catch up, if they could catch up at all. Google's Search Console data from those firms tells the story clearly: organic traffic plateaued while competitors' grew. Some are still paying for PPC because their competitors locked down the organic positions a decade ago.
AI search is following the same adoption curve, but faster. Much faster. SEO had a five-year early mover window. AI search has maybe two years. Maybe less. And as voice search adoption among legal consumers accelerates alongside AI, the firms not optimizing for conversational queries are losing ground on two fronts simultaneously.
We ran our full AEO/GEO playbook for Scarsdale Solicitors. The results: 4,130 clicks and 744,000 impressions in three months. That growth came from the compound effect of traditional SEO and AI optimization working together. You can see the full breakdown on our case studies page.
The question isn't whether your firm should optimize for AI search. It's whether you'll be one of the firms that moves now, or one that spends three years and three times the budget trying to catch up. Check out our services page to see how we're building AI search optimization into every campaign, or run a free SEO audit to see where you stand today.
Here's something most AI marketing guides won't tell you: bar advertising rules don't stop at your website. If AI surfaces claims about your firm that violate your state's advertising ethics rules, you could face a bar complaint, even if you didn't write the AI's response.
The good news: the same content that performs best with AI is also the most ethically safe. Factual statements, verifiable data, neutral tone. That's exactly what bar ethics rules want. It's the superlative marketing language ("best lawyer," "guaranteed results," "we never lose") that creates both ethical risk and poor AI performance. So the incentives are aligned.
But there are specific considerations law firms need to watch for:
AI hallucinations. AI models sometimes generate inaccurate information about firms: wrong practice areas, incorrect locations, fabricated case results. You can't prevent this entirely, but you can reduce it by providing clear, structured, consistent information across your web presence. The more factual sources an AI model can draw from, the less likely it is to fill gaps with fabricated details. Monitor what AI platforms say about your firm quarterly and correct source content when inaccuracies appear.
Unverifiable claims. If your Brand Facts page states a 94% success rate, that number needs to be accurate and defensible. AI models may surface it in responses, and other attorneys (or bar associations) may see it. Don't include statistics you can't verify. Don't claim specializations your jurisdiction doesn't certify.
Testimonial and review content. AI models sometimes pull from client reviews and testimonials when generating firm descriptions. Ensure your review responses and testimonial pages comply with your state's bar rules. Some jurisdictions restrict how client outcomes can be discussed in marketing materials, and AI doesn't distinguish between a testimonial page and a practice area page.
For a deeper treatment of the compliance angle, our AI chatbot compliance and ethics guide covers the full regulatory picture. The firms doing this right aren't choosing between performance and compliance. They're finding that the two are almost perfectly aligned. If you need help navigating both the optimization and compliance sides, our guide to choosing a law firm SEO agency covers what to look for in a partner who understands legal advertising rules.
Continue reading
Full breakdown of keyword research, local SEO, link building, and technical optimization.
Read guideGoogle Maps, GBP optimization, citations, and review strategy.
Read guidePricing data, ROI benchmarks, and budget planning for legal SEO.
Read guideFrequently asked questions
01
AEO (Answer Engine Optimization) focuses on getting your firm cited as the answer when someone asks an AI assistant a question. Traditional SEO targets search engine result pages. AEO targets the answer that ChatGPT, Perplexity, or Google AI Overviews deliver directly. The two overlap significantly but AEO requires more structured content and authoritative sourcing.
02
Ask ChatGPT, Perplexity, and Google (with AI Overviews enabled) questions your potential clients would ask, like "best personal injury lawyer in [city]" or "how to file for divorce in [state]." Note whether your firm is mentioned or cited. Check if your content appears in Perplexity's source citations. This is a manual process today with no comprehensive tracking tool yet.
03
Approximately 41% of legal searches now interact with AI in some form, including Google AI Overviews, ChatGPT, Perplexity, and Bing Chat. This number is expected to exceed 60% within the next 18 months as AI tools become the default interface for more search platforms.
04
AEO (Answer Engine Optimization) focuses on getting cited when AI answers a direct question. GEO (Generative Engine Optimization) is broader: it covers how your content performs when AI generates comprehensive responses, summaries, or recommendations. GEO includes brand entity optimization, structured data, and authority signals that influence how generative models perceive your firm.
05
Yes. AI platforms pull from indexed web content, so strong SEO is the foundation for AI visibility. Firms that rank well organically are more likely to be cited by AI tools. The shift is not away from SEO but toward SEO plus AEO plus GEO as a combined strategy.
06
ChatGPT draws from its training data and web search results (when connected). It favors firms with: authoritative content that directly answers legal questions, strong third-party mentions (directories, news, legal publications), consistent structured data, and a clear topical footprint around specific practice areas.
07
A Brand Facts page is a structured summary of your firm's key information: founding year, practice areas, notable results, attorney credentials, office locations, and contact details. It gives AI platforms a clean, authoritative source to pull from when generating responses about your firm. Think of it as a Wikipedia-style fact sheet on your own domain.
08
Yes. Structured data gives AI platforms machine-readable facts about your firm: services offered, locations served, attorney qualifications, and client reviews. LegalService and Attorney schema types are particularly useful because they map directly to the kinds of questions people ask AI about lawyers.
09
Bar advertising rules apply to any channel where potential clients find your firm, including AI. Avoid making claims through AI optimization that you could not make in a traditional ad. Do not fabricate credentials, outcomes, or specializations to influence AI responses. The same ethical standards apply regardless of the medium.
10
Your competitors who optimize for AI will be cited and recommended while your firm will not. As the percentage of AI-mediated searches grows past 60%, firms without AI visibility will see declining lead volume even if their traditional SEO rankings hold. The early mover window is approximately two years from now.
Next step
We will run an AI visibility audit across ChatGPT, Perplexity, and Google AI Overviews for your practice areas and markets.