Key takeaways
- Legal queries trigger Google AI Overviews roughly 78% of the time, the highest rate of any category SE Ranking has tracked, so the stakes for showing up in AI answers are higher for law firms than almost any other industry.
- A small group of directories (Chambers, Legal 500, Super Lawyers, Best Lawyers, Martindale, Avvo, Justia) currently dominates AI citations for legal queries, sometimes beating a firm's own practice page for searches about that exact firm.
- Reddit's share of ChatGPT citations collapsed from about 3.8% to 0.5% in a single week in August 2026, so advice to "build a Reddit presence" for AI visibility needs a serious update.
- Dedicated AI visibility platforms range from about $95 to $500+ a month; general SEO suites like Semrush bolt on AI tracking as an add-on; a few legal-specific tools bundle monitoring with directory and schema fixes.
- Tracking is only half the job. The firms making progress are pairing monitoring with content fixes, structured data, and consistent directory profiles, not just watching a dashboard.
Why law firms can't ignore this anymore
I'll be blunt about the number that matters most here: legal searches trigger Google AI Overviews about 78% of the time, according to SE Ranking research cited by the American Bar Association. That's the highest rate of any vertical SE Ranking tracked. If you run a law firm and you're not watching what AI Overviews, AI Mode, ChatGPT, and Perplexity say about you, you're ignoring the channel where most of your prospective clients' first research happens.
And it gets stranger. The same ABA research, done with Near Media, found that when people were asked where they went to learn about the legal issues in their case, 54% said "Google or another search engine" and only 13% said an AI tool or chatbot. Fewer than 2% said they'd use AI to find a lawyer directly. So the real pattern looks less like "AI replaces the firm's website" and more like "prospect researches on AI, then does a branded Google search to find your phone number." That shows up in your analytics as rising impressions for your firm's name with flat click growth, which is exactly the kind of signal a generic rank tracker won't explain to you.
There's also a real downside risk that doesn't exist in most other industries. Researchers at the Charlotin database tracked 487 documented AI hallucination cases in US court filings in 2025, ten times the 2024 total, and 37.8% of those involved licensed attorneys. Separately, AI tools have a habit of turning a vague bio line like "handles some employment matters" into an unearned "employment law specialist" claim, which can run into bar rules on unsubstantiated specialization claims. Visibility tracking for a law firm isn't just "are we mentioned"; it's "are we being described accurately and in a way that won't get us a bar complaint."
What AI visibility actually means for a law firm
Most generic SEO tools answer one question: did we rank for this keyword. AI visibility tools try to answer a different, messier set of questions:
- When someone asks ChatGPT, Perplexity, or Google AI Mode something like "best personal injury lawyer in Denver" or "do I need a lawyer for a car accident in Texas," does your firm get named?
- When it does get named, where is the AI pulling that information from? Your own site, a directory profile, a news mention, a bar association page?
- How does the AI describe your firm, and is that description accurate and compliant?
- Are competitors showing up instead of you, and why?
This matters because the citation mechanics differ a lot by engine. Promptwatch's own data on average sources per response shows ChatGPT typically cites around five sources per web-search answer, roughly half of a traditional Google results page, while Google AI Overviews cites about ten and Perplexity sits right around ten as well, consistently, day after day. Microsoft Copilot is the wild card: its average citation count has swung from under two to nearly seventeen per response within weeks, which tells you Microsoft is still rebuilding how Copilot attributes sources. If you're a law firm trying to get cited, ChatGPT is simply a tighter race for fewer slots than AI Overviews or Perplexity.
There's good news buried in there too. Promptwatch's domain rank breakdown for ChatGPT citations from August 2026 shows domains in the DR 46-75 range picking up nearly half of all citations, while the biggest, most authoritative sites (DR 91-100) actually fell to around 3% of citations by mid-August. A mid-sized firm with a solid, well-structured site doesn't need Am Law 100 domain authority to get picked up by ChatGPT. That's a different story than classic SEO, where domain authority alone can feel like destiny.
The directory problem is bigger than most firms realize
Here's the part that should change your strategy, not just your tooling. Research from 5WPR's 2026 Legal AI Visibility Index found that a small cluster of directories, Chambers, Legal 500, Super Lawyers, Best Lawyers, Martindale, Avvo, and Justia, dominates the AI citation layer for nearly every type of legal query tested. In some cases these directories outrank an elite firm's own practice page even when the query is specifically about that firm. The example cited: Cravath, Swaine & Moore's own M&A page lost out to directory listings for queries about Cravath's own M&A practice.
That's a strange thing to sit with. Your firm's own website, written by people who know the firm best, can lose a visibility fight to a third-party profile page. The 5WPR report calls this "recursive citation": once an AI engine starts naming a firm for a category, it tends to keep naming that firm, which compounds an advantage for firms that got cited early. If you haven't been actively managing your directory profiles, you're behind firms that have, and the report warns the cost of catching up on citation authority could rise 50-80% compounded annually over the next two years. Whether or not that exact figure holds, the direction is believable: this is a window, not a permanent state.
A caveat on forum advice: Reddit is not what it was
A lot of legal marketing advice still tells firms to build presence on Reddit, Quora, and Avvo forums because AI models cite forum content. That advice needs an asterisk now. Promptwatch's tracking shows Reddit's share of ChatGPT Search citations fell from a steady ~3.8% (mid-July to early August 2026) to about 0.5% on August 14, 2026, an 86% relative drop in about a week. You can see the full breakdown in Promptwatch's data on Reddit citations dropping in ChatGPT. Google AI Overviews and AI Mode are declining too, just more gradually, AI Overviews down about 11% and AI Mode down about 30% over the same window.
So: Reddit still matters more for Google's AI surfaces than for ChatGPT right now, and that balance could flip again next month. This is exactly the kind of thing a point-in-time audit misses and ongoing monitoring catches.
What content actually gets cited
Promptwatch's ChatGPT citation type data for August 2026 is useful here: product pages led at 28.7%, listicles at 10.1%, how-to content at 6.3%, news articles at 5.6%, and comparisons at 3.5%. The bigger story is the trend inside the month: how-to content more than doubled, from about 4.3% in the first week to 9.1% in the last nine days, while social posts collapsed from 4.4% to under 1%. For a law firm, that's a vote in favor of "what to do after a car accident" or "how to file an unemployment claim in Ohio" style guides over anything that looks like a forum post or social update.
A Google generative-search guide decoded by Just Legal Marketing in May 2026 adds a technical detail worth knowing: AI Overviews and AI Mode use the same crawling, indexing, and ranking infrastructure as regular Google Search. There's no separate AI index and no AI-specific schema type, and llms.txt files don't improve citation odds (content chunking for AI can even hurt authority signals in some cases). What reportedly does help: standard schema (LegalService, Person, FAQPage, LocalBusiness), consistent NAP data across every listing, and short, direct "atomic answer" paragraphs of 40-60 words right under each heading.
Family law content is a good example of this working in practice. Research from Lexicon Legal Content found pages with 8-12 natural, emotionally worded FAQs, each paired with a 40-80 word direct answer and FAQPage schema, earn noticeably higher AI citation rates than generic FAQ pages, with custody questions generating the most AI Overview citations within the practice area.
One more technical check that's easy to overlook: confirm your developer hasn't blocked ChatGPT-User, PerplexityBot, ClaudeBot, OAI-SearchBot, or Google-Extended in robots.txt "as a precaution." It happens more than you'd think, and it silently removes a firm from every AI platform at once. The fix takes minutes once you know to look.
Comparing AI visibility tools for legal services
| Tool | Starting price | Engines covered | Best fit for a law firm |
|---|---|---|---|
| Promptwatch | $95/mo (Essential) | ChatGPT, Gemini, Claude, Perplexity, Grok, AI Overviews, AI Mode, and more | Firms that want monitoring plus content generation, crawler logs, and CMS publishing in one place |
| Mentionova | $99/mo | ChatGPT, Perplexity, Claude, Gemini, AI Overviews, Reddit (full 6-engine coverage on Scale+) | Firms wanting built-in defensive and comparison prompt tracking ("Firm A vs Firm B") |
| Semrush AI Visibility Toolkit | $99/mo add-on (requires a core Semrush plan) | ChatGPT, Google AI, Gemini, Perplexity | Firms already paying for Semrush who want a bolt-on AI layer |
| Profound | ~$99-499/mo (estimates vary by source) | ChatGPT at entry tier, up to multiple engines on Growth | Larger firms or legal departments needing multi-region tracking |
| Peec AI | ~$95-495/mo | 3 to 6 engines depending on tier | EU-based or cross-border firms needing GDPR-aligned reporting |
| Trakkr | $100/mo | All 8 models on every paid plan, no per-engine fees | Legal-tech vendors or firms wanting Reddit and crawler intelligence bundled in at every tier |
| TendorAI | Free report / £299/mo Pro | UK-focused | UK solicitors who want schema and directory-listing fixes included, not just monitoring |
Worth flagging: the Rankability review of Semrush's AI Toolkit points out its data comes from simulated prompts, not live user sessions, so treat any vendor's "AI market share" number as a directional trend rather than a precise audit. For high-stakes decisions, manually run the prompts yourself in ChatGPT or Perplexity and compare.
Where Promptwatch fits for law firms
Most of the tools above answer "were we mentioned." Promptwatch tries to answer the next question too: why aren't we mentioned, and what do we do about it. For a law firm that doesn't have a dedicated GEO team, that's the gap that matters. Its crawler logs show exactly when ChatGPT-User, PerplexityBot, or ClaudeBot actually visit a practice area page and whether they hit an error, which is more useful than a monthly citation score when you're trying to diagnose why a specific page never gets cited.
Promptwatch also tracks Reddit and YouTube citations directly, which matters given how fast Reddit's share is moving right now, and it includes a content gap analysis and Content Agent that can draft and publish AEO-optimized practice area pages to a CMS like WordPress or Webflow. For a firm managing dozens of practice-area and location pages, the alternative is a partner or marketing coordinator manually checking ChatGPT every week, which doesn't scale once you're tracking 50+ prompts across multiple practice areas and cities.
Pricing starts at $95/month for the Essential plan (50 prompts, 6,000 responses, 5 AI-optimized articles, all major LLMs tracked), scaling to $245/month for Professional, which adds state and city-level tracking, a feature that matters a lot for firms competing in specific metros rather than nationally.
A short list of other tools worth knowing about
A few other names show up repeatedly in legal visibility comparisons and are worth a quick look if you want to shop around. SE Ranking's AI visibility toolkit is a reasonable option if you're already on their platform for traditional rank tracking.
Ahrefs Brand Radar leans more toward legal-tech vendors and SaaS companies than firms themselves, but it's worth a look if you're already deep in the Ahrefs ecosystem for backlink work.
For firms running an agency relationship, Scrunch AI and Otterly.AI are both commonly bundled into broader digital marketing retainers as a lighter-weight monitoring layer.
Ethics and compliance: the part most GEO vendors skip
This is the section a lot of marketing writeups leave out, and for a law firm it's not optional. The State Bar of Arizona's formal guidance on AI tools maps cleanly onto AI visibility work: competence (you need to actually understand how the tool generates its claims, not just trust the dashboard), diligence (don't copy-paste AI-drafted content without a lawyer reviewing it), confidentiality (never feed client information into a public AI visibility or content tool), candor (verify anything AI-generated before it goes on your site, because hallucinations happen), and misconduct (don't let a tool's auto-generated bio copy overstate a practice area into something that looks like an unearned specialization claim).
The ABA's own coverage of this space is blunt about one more thing: be skeptical of any vendor promising guaranteed LLM rankings. Nobody, including the people who built the models, fully understands retrieval behavior yet. A tool that promises a number one spot in ChatGPT is selling you something it can't actually deliver.
A practical starting checklist
- Audit your presence on the seven dominant legal directories (Chambers, Legal 500, Super Lawyers, Best Lawyers, Martindale, Avvo, Justia) before worrying about anything else; this is where most legal AI citations originate.
- Check robots.txt on your own site for accidental blocks on AI crawlers. It's a five-minute check that can explain a mysterious invisibility problem.
- Add FAQPage, LegalService, and Person schema to your top practice-area pages, and make sure your NAP data is identical across every listing.
- Consolidate thin, duplicate practice-area pages into one authoritative, regularly updated page per topic; splitting the same topic across several pages dilutes citation eligibility.
- Pick one monitoring tool, run it for a full quarter across at least ChatGPT, Perplexity, and Google AI Overviews, and track whether branded search impressions rise even if clicks stay flat, that's often the real signal of progress.
For firms that want to see what a fuller tool landscape beyond legal-specific options looks like, the GEO software directory at bestgeosoftware.com is a good place to browse by feature set rather than vertical, and agenticseotools.com is worth a look if you want to see which platforms, like Promptwatch, go beyond monitoring into automated content and fixes.