How to Get Your Law Firm in ChatGPT and Google AI Overviews (2026)

So I got a call maybe three weeks ago from a managing partner who said his traffic was down 30% and he couldn’t figure out why because his rankings hadn’t changed and his ad budget was the same and everything in the monthly report looked normal. And when I dug into it what I found was that Google had started putting AI written answers above his organic listings on about half the searches his firm used to own, and people were reading those answers and leaving without ever clicking his website. His content wasn’t ranking worse. It was being read by Google and summarized for free at the top of the page, and the searcher never made it down to his listing. The industry calls it zero-click cannibalization and it’s the single biggest shift in legal marketing since Google started selling ads.

I should say upfront that most of what gets published about AI and law firm marketing comes from agencies selling AI optimization services. This page is built on three sources that aren’t selling anything; a peer reviewed study from Princeton, a Harvard Journal on Legislation and Technology commentary that analyzed how AI treats law firm content, and a Pew Research study of nearly 70,000 real Google searches. I’ll tell you where the evidence is strong and where honestly nobody knows yet, and I’ll cover the part that’s making a lot of managing partners nervous right now; how to get your content cited in AI answers without letting AI bots scrape your proprietary strategies to train models that could end up helping your competitors.

AI SEARCH IMPACT ON LAW FIRM TRAFFIC: ZERO-CLICK DATA AND CITATION ECONOMICS

---

The Pew study found that when Google puts an AI answer at the top, people click regular search results about half as often as they normally would. The click rate on the number one organic result drops by 58% and over a quarter of searchers just read the AI summary and leave without visiting any website, which if you’re paying $200 a click for injury keywords means Google is giving away for free what you paid to generate. And the decline is even worse than that headline number suggests; for informational keywords like “what are the penalties for a first DUI” the click rate on the top result has gone from roughly 7% in 2023 to about 1.6% now, which means “page one” for those queries doesn’t just mean less traffic anymore; it means almost zero traffic unless you’re the source the AI is actually citing.

But and this is the part that made me rethink everything; the AI answer doesn’t replace your listing if you’re the cited source; it actually amplifies it because the searcher sees your firm’s name in the summary and trusts it more and then clicks your result below. The firms getting cited are winning twice and the firms not getting cited are losing queries they used to own, which honestly changes the entire math on content strategy for law firms. I should be clear about what “winning” means here though; Pew found that only about 1% of people actually click the links inside the AI summary itself. The value isn’t direct referral traffic from the citation. The value is that the searcher sees your firm’s name in the AI answer, trusts it, and either clicks your regular listing below or Googles your firm name later to verify. It’s a brand impression, not a click, and you have to measure it that way.

And there’s a piece of good news I don’t see anyone talking about; local searches like “lawyer near me” or “personal injury attorney Miami” trigger AI summaries less than 8% of the time. The queries getting disrupted are mostly informational; “what are the penalties for a first DUI” and “how long do I have to file a lawsuit after a car accident.” So if your practice lives on local search and the map results, AI hasn’t really touched your main traffic source yet, but if you depend on blog posts and educational content to bring people to the site then the disruption is already here and it’s accelerating.

HOW TO GET YOUR LAW FIRM CITED IN AI ANSWERS: ANSWER CAPSULES AND E-E-A-T SIGNALS

---

So the Harvard JOLT commentary analyzed how AI systems evaluate law firm content and the finding that stuck with me is this; pages that stated the actual legal rule or process in the first two sentences got cited in AI answers and pages that opened with “Facing a DUI? You need an experienced attorney” were consistently skipped, even when the real legal information was accurate further down the page. And I think about this all the time now because maybe 90% of the practice area pages I’ve ever audited open with a sales pitch, not an answer, and apparently the AI just reads that sales pitch and moves on to the next site that gave it something useful.

The other findings were honestly just as bad for most firms. Pages using actual questions as section headings performed better than pages with generic headings like “Our Approach” or “DUI Defense.” Attorney bylines with credentials were mostly missing, which matters because AI systems use author credentials as a trust signal. And local landing pages that were basically just a city name and a phone number rarely got cited; the pages with actual local court details and specific statutory references were the ones getting pulled into AI answers.

Here’s what the difference looks like on an actual practice area page.

The first version is a sales pitch and AI recognizes it as one. The second version states the law, cites the statute, gives actual numbers, and answers what the person searched for. Over 70% of pages generating traffic from ChatGPT opened with this kind of direct answer at the top; they call it an “answer capsule” and it’s probably the single most useful thing I’ve learned about AI search in the last year. And here’s the part that changes your website strategy; if the AI already gave the searcher the legal answer before they clicked your site, they don’t need you to explain the law again when they land on your page. They already know the statute of limitations is four years. What they need now is proof that you’re the right firm to handle it; case results, reviews, attorney credentials, and a phone number. The education phase happened on Google. Your site’s job is the trust phase.

GOOGLE AI OVERVIEWS VS CHATGPT FOR LAWYERS: TWO DIFFERENT SYSTEMS, TWO DIFFERENT STRATEGIES

---

This is the part I don’t see anyone else explaining clearly and it matters because most firms treat “AI search” as one thing when it’s actually two completely different systems that pull from different sources and require different strategies to show up in.

And here’s the part that really changes things; 90% of the web pages ChatGPT cites rank at position 21 or lower in Google. You don’t need to be in Google’s top 3 to get recommended by ChatGPT. What you need is content that contains specific, verifiable answers the AI can’t find on every other law firm website, which is a completely different game than traditional keyword rankings. The firms that win in both systems are the ones rewriting their practice area pages to open with direct legal answers (the Google AI play) while simultaneously building out their directory profiles and collecting reviews across multiple platforms (the ChatGPT play). Neither strategy alone is enough anymore.

LAW FIRM ENTITY SEO: WHY AI RECOMMENDS SOME FIRMS AND SKIPS OTHERS

---

AI doesn’t rank websites. It recommends entities. And an “entity” is just a fancy way of saying a verified thing that exists across multiple trusted data sources; a specific attorney, a law firm, a practice area, a location, and the connections between them. When someone asks “best car accident lawyer in Miami” the AI doesn’t scan websites for keywords; it queries its internal database to find entries tagged as “Attorney” plus “Personal Injury” plus “Miami” and cross references their trust scores. If your firm doesn’t exist as a clearly defined entity in that database, you’re invisible regardless of how good your website is.

The trust score comes from consistency. If your website says “123 Main Street” but your Google Business Profile says “123 Main St.” and your Yelp says “125 Main Street” the AI sees three different entities instead of one strong one. It drops your score because conflicting data means uncertainty and uncertainty means the AI would rather show nothing than risk recommending the wrong firm. And the name match matters even more than the address; if your website says “Jon Doe” but the State Bar says “Jonathan Q. Doe” the AI sees a weak entity link and might skip you entirely no matter how good your content is, because it can’t confirm you’re actually a licensed attorney. I’ve seen this kill visibility for firms that moved offices and updated Google but forgot about the 15 other directories that still had the old address, and I’ve seen it happen with attorneys who use a nickname on their website instead of their bar registered legal name.

And reviews are read now, not just counted. The AI performs what’s called aspect based sentiment analysis, which is a complicated way of saying it reads the actual text and extracts specific attributes. A review that says “Great lawyer!” provides almost no data. A review that says “they kept me informed about my custody case every week and explained the process clearly” gives the AI multiple data points linking your firm to a case type and specific service qualities. The rate of new reviews matters too; 100 reviews from 2022 is less valuable than 50 from 2025 because review velocity signals an active practice. And when you reply to a review, that text also gets indexed, so use responses to reinforce what you do; “we were glad to help you resolve your complex probate matter involving out of state assets” is doing double duty. And the flip side of this is something nobody talks about; the AI also reads negative reviews and can surface old disciplinary actions or bar complaints if those are prominent in your data. This isn’t just about building positive visibility; it’s about making sure the negative signals in your profile aren’t louder than the positive ones, because AI doesn’t have judgment about context the way a person does. A resolved complaint from 2018 can still show up in a ChatGPT answer in 2026 if nobody bothered to bury it with newer, stronger signals.

ROBOTS.TXT, SCHEMA MARKUP, AND INDEXNOW: TECHNICAL SETUP FOR AI SEARCH VISIBILITY

---

There are three technical changes that determine whether AI systems can even see your firm, and the first one takes about 10 minutes and most law firm websites have it configured wrong.

First, your robots.txt file. OpenAI runs three separate crawlers and each one does something different. OAI SearchBot is the one you must allow because it surfaces your pages in ChatGPT’s live search answers and it does not scrape content for model training. GPTBot is the training crawler you can block if you want to protect proprietary content. And ChatGPT User is the one that fires when someone asks ChatGPT to summarize a specific URL; as of December 2025 OpenAI declared it ignores robots.txt entirely because it acts as a proxy for human browsing, so you can’t block that one even if you wanted to.

Second, Schema Markup. Your site needs invisible code underneath the content that tells AI systems what your firm does, who your attorneys are, what practice areas they cover, and what questions your pages answer. The correct structure uses LegalService as the parent type for the firm and Attorney as a subtype of Person for individual practitioner pages. The property that matters most is called “sameAs” and it’s basically a list of URLs that all refer to the same entity; your State Bar profile, your Avvo page, your Justia profile, your LinkedIn. This tells the AI “the entity on this page is the exact same entity verified at these trusted locations” and it merges the trust scores together. And while you’re at it, ask your developer about creating an llms.txt file in your site’s root directory; it’s a plain text file that gives language models a human readable summary of your firm’s practice areas, location, and credentials so AI systems that don’t use traditional crawlers can still understand who you are. The SEO guide covers the full technical implementation.

Third, IndexNow for Bing. Since ChatGPT pulls from Bing in real time, the difference between your content being indexed in 30 seconds versus 3 weeks is the difference between being cited in today’s answers or being invisible until next month. IndexNow is a protocol that pings Bing the moment you publish or update a page instead of waiting for Bing’s crawler to find it on its own. Most WordPress plugins support it now and the setup takes maybe an hour. In a world where ChatGPT’s live search depends on what Bing has indexed right now, this is not optional anymore.

And there’s a fourth thing that isn’t technical but might matter more than any of them; YouTube. Google’s AI Overviews increasingly pull answers from YouTube transcripts, and a 60 second video of the attorney explaining the legal process in plain language does more for AI visibility than 2,000 words of text on the same topic. If your practice area page explains the statute of limitations, record a video of the attorney saying the same thing and embed it on that page. The AI now has two sources from the same entity saying the same information, which reinforces the trust score, and it can deep link to the specific moment in the video where you answer the question. Despite this, only about a third of law firms create any video content at all, which is one of those gaps where doing the thing almost nobody else is doing gives you a disproportionate advantage.

TRACKING AI SEARCH REFERRALS: TOOLS, PROXY METRICS, AND WHAT ANALYTICS CAN’T SHOW YET

---

And this is the part that drives everyone crazy; Google doesn’t tell you which traffic came from AI Overviews and ChatGPT doesn’t pass clean referral data when someone clicks a citation link. So you can’t just look at your analytics dashboard and see “AI search sent you 47 visits this month” the way you can with Google organic. About 65% of all Google searches now end without a click, and for queries triggering AI Overviews specifically that number is closer to 83%. Your website traffic stays flat but your phone rings more and you have no idea why; that’s the tracking gap in one sentence.

And there’s a deeper problem that SparkToro’s research uncovered; AI systems rarely give the same recommendation twice for the same query. Ask ChatGPT “best personal injury lawyer in Miami” three times and you might get three different answers depending on the time of day, what it’s already discussed in your conversation, and how its retrieval system happens to sample Bing’s results that moment. Which means every visibility percentage I just cited is a moving target, not a guaranteed outcome, and anyone telling you they can “guarantee” ChatGPT placement is making promises the technology itself can’t keep.

The proxy metric that works surprisingly well is watching for unexplained increases in direct website traffic or branded search volume, because the most common behavior is someone reads a ChatGPT recommendation, skips the citation link entirely, and just Googles the firm’s name to verify what the AI told them. If your Bing Webmaster Tools impressions are going up on practice area queries and your direct traffic is going up at the same time, AI is probably sending you business even though you can’t see it in your analytics, which I know is a frustrating answer but it’s the honest one.

AI SEO FOR LAW FIRMS: WHAT THE RESEARCH SUPPORTS AND WHAT’S STILL UNPROVEN

---

Agencies now charge up to $8,000 a month specifically for AI search optimization, and the industry publication Attorney at Law Magazine reported that this comes “with largely unproven ROI.” Nobody has actually measured whether leads from AI recommendations convert to signed cases at a higher or lower rate than regular search leads and I don’t think anyone will for a while because the tracking is still messy. The rules change constantly and what works now probably won’t work exactly the same way in six months.

But the data also shows that mentions of your firm on other websites have the strongest correlation with appearing in AI results; stronger than anything you do on your own site. Which means the most effective AI visibility strategy might be the cheapest one; getting mentioned and linked from bar associations, legal publications, and local news coverage. And the firms moving on this now probably have maybe 12 to 18 months before everyone else catches up, which in my experience is a window worth taking seriously even if the specific numbers keep shifting along the way.

Related: SEO for Lawyers · The 5 Channel Growth Engine · Conversion Optimization · Complete Strategy Guide · PI Marketing Services · ROI Benchmarks