Which AI Engines Actually Matter for Your Law Firm’s Visibility

Graphic with teal globe icon and text "Which AI engines matter when it comes to your firm's visibility?

If you’re an attorney, you’ve heard all of it by now. Get found in AI. Set up your website for AI. Get cited in AI results.

So which ones actually matter for your firm?

Short answer: two of them. The rest are noise. And one of the two probably isn’t the one you’re thinking of.

Here’s how we got here. In early 2024, Gartner predicted that traditional search engine volume would fall 25% by 2026 as people moved to AI chatbots. Every marketing agency in the legal space repeated it, including some of the people reading this. It’s 2026. Here’s what actually happened.

    What Can We Help You With Today?

    (Check all that apply)


    What is your monthly marketing budget?

    When are you looking to get started?

    Can you provide your contact details?




    Disclaimer: By submitting your phone number, you consent to receive SMS messages from Civille at the number provided. Message frequency may vary and may include updates or offers related to your inquiry. Message and data rates may apply. View our Privacy Policy and Terms of Use.

    What’s your name?



    Thanks [first-name]! What are your business details?




    The Prediction Everyone Repeated Was Wrong


    Search volume didn’t drop 25%. Alphabet reported Q2 2026 Search revenue of $63.3 billion, up 17% year over year, and Sundar Pichai told investors that AI features are driving an incremental increase in query volume, not a decrease. Google still accounts for roughly 88% of all search referrals across Cloudflare’s network as of July 2026.

    So the forecast missed. But something else broke too, and it’s the thing that’s actually costing law firms traffic.

    Clicks collapsed. SparkToro’s clickstream panel found that 68% of Google searches ended without a single click between January and April 2026, up from 60% in 2024 and 49% in 2019. Ahrefs, tracking 75,000+ sites, measured a roughly 22% decline in search traffic between June 2025 and May 2026.

    Here’s the thing, though – people are actually searching more. But they’re clicking less. That’s the story from my post last month’s on traffic loss, and it’s the foundation for this one. If you’re deciding where to put your money based on “AI is replacing Google,” you’re solving the wrong problem.

    Google Won the AI Transition

    This is the part most legal marketing content gets backward.

    The assumption is that ChatGPT is pulling users away from Google. What the data shows is Google absorbing the AI opportunity almost entirely. Similarweb’s tracking of AI-assistant traffic tells the story in one line:

    AssistantShare of AI Assistant Traffic (June 2025)Share, May 2026
    ChatGPT~76%~54%
    Geminiunder 9%~28%
    Claude~2%~9%
    Perplexity~2%~1.3%


    ChatGPT lost about 23 points of share in eleven months. Nearly every point of it went to Gemini.

    Then there’s the part that never shows up in a referral report. Google’s AI Mode passed 1 billion monthly active users as of the Q2 2026 earnings call. The Gemini app crossed 1 billion monthly active users separately in August 2026. AI Overviews now appear on roughly 43% of Google searches, up from about 15% a year earlier.

    When someone uses AI Mode or reads an AI Overview and then clicks through to your site, your analytics logs that as Google organic. Not as AI. Every study measuring “AI referral traffic” undercounts Google by a wide margin, and Previsible said so directly in its July 2026 report: Google’s AI surfaces almost certainly drive more traffic than all standalone chatbots combined.
    eMarketer projects Google’s unduplicated AI user base will surpass ChatGPT’s by Q1 2027 and continue to extend its lead thereafter.

    For a law firm, “which AI engine should I optimize for” is still mostly a Google question. That’s good news, because the work is work you already know how to do.

    Which Engines Your Potential Clients Are Actually Using


    Reach and influence aren’t the same thing, so it helps to look at both.

    Pew Research surveyed 5,119 US adults in February 2026 and found 49% now use AI chatbots, up from 33% in 2024. By platform reach: ChatGPT 44%, Gemini 24%, Copilot 17%, Meta AI 14%, Grok 8%, Claude 6%. The top use case was information searching at 42%. And 60% of Americans say they read the AI summaries that appear in search results.

    Notice the gap between reach and engagement. Copilot reaches 17% of US adults but represents about 1.3% of assistant traffic, because it’s bundled into Windows and Office rather than chosen. Same pattern with Meta AI.

    Now the legal-specific piece. iLawyer Marketing surveyed 1,110 consumers in July 2026 about which online sources they’d use to research an attorney for an important legal issue:

    • Google: 71.9%, down from 86.7% in 2025
    • ChatGPT: 41.9%, up from 28.1%
    • Any AI answer engine: 50.1%
    • AI only, no traditional source: 9.5%
    • Yelp fell to 13.6%, Reddit to 11.7%

    ChatGPT’s trend line for attorney research runs 9% in 2023, 21% in 2024, 28% in 2025, and 41.9% in 2026. Worth noting this is an agency-run survey asking what people would use rather than tracking what they did, so treat it as directional. But it lines up with everything else.

    According to a study by BrightLocal in March 2026, 45% of consumers have used AI to find local business recommendations, compared to only 6% a year ago. AI is now the third-most-used discovery tool behind Google and Facebook. “ChatGPT was the leader with 31%, followed by Google AI Mode with 23%.

    Two things in this data surprise most attorneys:

    The age skew runs the opposite direction from what firms expect. In the same the legal survey, 45-to 60-year-olds were the heaviest AI users for attorney research: 57% used ChatGPT, 76% used some AI tool. Ages 18 to 29 came in at 35% and 48%. BrightLocal found the same inversion among local businesses, with the 30- to 44-year-old group at 64% and the over-60 group at 24%. The people using AI to find a lawyer are the people with the money and the legal problems.

    Claude and Perplexity are close to noise for consumer legal research. Claude’s traffic share tripled this year, but according to Anthropic’s own data and economic index, most of the use is occupational. Majority of Claude users show “computer and mathematical” occupations, with over 20% of conversations involving Claude Code. Pew data puts Claude’s consumer reach at 6%, the lowest among named AI AI assistants. Perplexity use amongst the legal space peaked in October 2025 and has been declining since. Most people use Claude for work (like me).  They use ChatGPT and Google/Gemini to find a lawyer or do legal research.

    Practical read: optimize for ChatGPT and Google’s AI surfaces. Everything else is a rounding error you get for free by doing those two well.

    AI Referral Traffic Is Still Small. It Also Converts Better Than Anything Else.


    Both of those things are true, and holding them at the same time is the whole strategy.

    Semrush analyzed billions of visits across 50,000+ websites and found AI traffic at 0.14% of total web traffic against 16.04% for organic search. Ahrefs reports AI search at 0.5% of its own site traffic. Orbit Media, looking at 97 B2B lead generation sites and 28.9 million sessions through June 2026, also landed at 0.5%.

    Nobody is getting a flood of AI referrals right now. Anyone telling you otherwise is selling something.

    Here’s the other half. When we spot-checked our own law firm accounts, AI referral traffic had the highest key event rate of any channel. Not marginally higher:

    • AI referrals: roughly 12% to 16% key event rate (form fill, phone call, or chat)
    • Organic search: roughly 2% to 4%
    • Paid search: roughly 5%

    That’s a 5x or better spread over organic. And it isn’t a Civille anomaly. Semrush measured AI visitors converting at 4.4x standard organic. Seer Interactive clocked ChatGPT referrals at 15.9% against 1.76% for Google organic. Orbit Media found AI traffic converting at 1.91% versus 0.5% for Google Search across its B2B panel. Ahrefs saw AI search deliver 0.5% of traffic and 12.1% of signups.

    Most of that published research covers software and ecommerce, not legal. What matters is that our legal accounts land inside the same range as datasets running thousands of domains. Different vertical, same mechanism.

    The mechanism is simple. Someone who arrives from ChatGPT has already completed the research. They’ve compared options, got a plain-English answer about their situation, and picked a name. They’re not browsing when they come to your site. They’re deciding. The visitor who used to land on your blog post at the top of the funnel now gets that answer inside the chat window and only clicks when they’re ready to talk to someone.

    Small channel. Very good visitors. Which is exactly why the answer is “invest in it,” not “move your budget there.”
    And the accounts posting the strongest numbers aren’t the average ones. They’re the firms with deep landing pages, real structured data, and deliberate AI visibility work behind them. Those accounts beat the published benchmarks instead of just matching them.

    Where AI Actually Gets Its Answers About Law Firms

    Two uncomfortable findings here, and both change what “AI optimization” should mean for a firm.

    First, AI engines rarely link to anyone in professional services. Similarweb tracked the share of citations in ChatGPT prompts. Citations were at 1.6% in June 2025 and rose to 6.8% in May 2026, which sounds like progress until you look at what it is actually citing and the category split. About 23% is travel and hospitality (people searching for vacations, restaurants, etc.). Professional services are under 4%.

    More than 96 out of 100 ChatGPT answers about legal topics cite nothing at all. There’s no link to earn in the legal space. So really, the question isn’t how to get clicked; it’s how to get named.

    Second, directories dominate the citations that do exist. An analysis of 18,900 citations across 540 high-intent legal queries found a legal directory was the first source cited 77.8% of the time. The most-cited individual firm site appeared 36 times out of 18,900. In 34.2% of answers where the AI named a firm, that firm had no page of its own among the cited sources. The model knew the firm from directory data.

    (Don’t take the exact percentages as universal across all AI engines – that study above was run on Perplexity specifically.) But the trend is similar in other studies, including another April 2026 index that found Chambers, Legal 500, Super Lawyers, Best Lawyers, Martindale, Avvo, and Justia ruling the roost in legal AI answers

    This puts us in an awkward position, because we spend a lot of time telling firms not to build their web presence on Justia or FindLaw. That position has not changed. But there is a difference to be made:

    A directory profile is a citation asset. A directory-built website is still a bad deal.

    Claiming your Avvo and Justia profiles, keeping your name, address, and phone number identical everywhere, and ensuring your practice areas and bar admissions are accurate is data hygiene. It takes an afternoon. Do it, because those records feed the models.

    Renting your website, your content, and your lead flow from a directory is a different decision entirely, and the AI citation data doesn’t make it any better. If anything, it makes owned assets more valuable because the firms being named alongside the directories are the ones with substantive pages that the models can read.

    What Still Drives the Actual Cases


    None of the above is a reason to move budget. Here’s what the money is still doing.

    Your Law Firm’s Google Business Profile. Your GBP is the most direct line between a search and a phone call. A survey from BrightLocal in February 2026 found that 71% of consumers use Google to find local businesses, 97% read reviews, and 74% prioritize reviews from the last three months. Most people’s standards have tightened quickly: 68% now require a minimum rating of 4.0, up from 55% a year ago, and 19% expect a review response the same day. Whitespark’s 2026 expert panel puts review signals at roughly 20% of local pack ranking weight and GBP signals at 19%.

    The AI angle makes GBP more important, not less. Of the 45% of consumers using AI for local recommendations, 97% verify what the AI told them against real reviews before acting. AI recommends. Your Google profile closes.

    Local Services Ads. A 2023 click test on the query “car accident lawyer” gave LSAs 42% of clicks, the local pack 29%, and organic 22%. It was a first-click test on a static results page rather than live clickstream data, so hold it loosely. The economics hold up better. Pareto Legal, working from $3.3 million in managed legal ad spend, found LSA leads cost more than Google Ads leads ($205 versus $95 CPL) but closed better, producing a lower cost per signed case ($2,485 versus $2,971) and 50.3% of all signed cases.

    One thing to watch: Google began migrating LSAs into Google Ads as Performance Max campaigns in August 2026, starting with home services categories. Legal isn’t in their first planned phase, but it’s coming.  This sounds intimidating but mostly consolidates management in the same view as your Google Ads, removes manual max-CPL bidding, and you’ll lose historical LSA reporting (which only goes back a couple months). So export your data before your category migrates.

    Organic search. Still the largest source of qualified traffic by an order of magnitude, and it’s the substrate everything else runs on. Ahrefs analyzed 146 million search results and found AI Overviews appearing on 23.6% of legal queries, but only 7.9% of local searches. They show up on 57.9% of question-format queries.

    Read that carefully, because it’s the most useful number in this post. The queries that trigger AI Overviews are informational: “how long do I have to file a claim in Wisconsin.” The queries that produce clients are local and transactional: “car accident lawyer near me.” AI is eating your blog traffic. It is not, yet, eating the searches that turn into signed cases.

    The Impression You Make After the Click

    Here’s the part that connects all of it. Better traffic quality raises the cost of a bad website.

    When someone arrives from ChatGPT having already compared three firms, they’re not there to learn. They’re there to confirm a decision. Your site has one job: don’t give them a reason to reconsider. Slow load, dated design, a bio page from 2019, a contact form that asks for eleven fields. Any of those undo the work that got them there.

    Then there’s intake, which is where most firms lose the case they already won.

    Clio’s secret shopper study had a third party contact 500 US law firms. Only 40% answered the phone. Only 33% responded to email. 48% couldn’t be reached by phone at all. Prospects who reached an actual person were nearly 8 times more likely to recommend the firm than those who hit voicemail.

    Consumers have gotten less patient since. CallRail’s September 2025 survey found 78% have abandoned a business after an unanswered call and 82% will simply call a competitor. Martindale-Avvo’s 2026 legal consumer research found 46% expect a response within four hours, and 47.6% name responsiveness as a critical factor in choosing an attorney.

    You can win the AI citation, win the click, and still lose to the firm that picks up the phone.

    What To Actually Do This Quarter

    1. Keep doing SEO. Structured, substantive pages are what both Google and the models read. Nothing about AI changes the fundamentals, and firms with deep practice area and location pages are the ones getting cited.
    2. Don’t Ignore Your Law Firm’s Google Business Profile. It sounds small, but continue updating your categories, hours, adding photos, and encouraging users to leave reviews. It’s what users see when AI refers them over. We do this for lawyers too.
    3. Claim and clean your directory profiles. Not just your GBP. Claim Avvo, Justia, Martindale, Super Lawyers, Bing Places, Apple Maps, etc.. Identical NAP everywhere. One afternoon, then leave it alone.
    4. Write for being quoted, not just ranked. Provide direct answers in the first sentence under each heading (especially if your heading is a question). Real specifics: jurisdictions, statutes, dollar figures, timelines. Generic content doesn’t get named.
    5. Segment AI referrals in your analytics. You can’t argue about a channel you aren’t measuring, and most firms have no idea what their AI traffic is worth.
    6. Fix intake before you spend another dollar on visibility. Answer rate, response time, after-hours coverage. Measure it this month.
    7. Leave the budget where the cases are. LSAs, GBP, and organic still produce the volume. AI visibility is an investment in the next 24 months, funded out of your content and technical SEO work, not out of what’s currently generating signed cases.

    Common Questions

    Which AI engine should my law firm optimize for?

    ChatGPT and Google AI overviews (AI Mode and AI Overviews). Around 42% of consumers seeking an attorney use ChatGPT, which makes up about 54% of all AI assistant traffic. Google’s Gemini app and AI Mode both reached one billion monthly users in 2026. Perplexity, Copilot, Grok and Claude are marginal for consumer legal searches, and doing well for the first two is generally good enough.

    Is AI search replacing Google for finding a lawyer?

    No. Google use for attorney research fell from 86.7% to 71.9% in one year, which is a real decline, but it’s still the dominant starting point by a wide margin. Only 9.5% of consumers used AI exclusively. What’s actually changed is that Google now answers more questions on the results page, so fewer searches produce a click.

    How much traffic will AI engines send my law firm?

    Very little today. AI represents about 0.14% to 0.5% of web traffic from large-scale studies. It’s about quality, not the quantity. AI referrals convert at 12% to 16% in our client accounts versus 2% to 4% for organic search.

    Do AI engines use my Google Business Profile?

    Often, though the underlying data sources vary by engine and change frequently. Research has found that ChatGPT surfaces Google Business Profile fields directly (such as hours, ratings, review counts, etc.) in local answers. Other studies cite Bing Places, Foursquare, and Yelp as business directories. Since you can’t control which source any given engine reads, keep them all accurate and consistent.

    Should I move budget from Google Ads to AI optimization?

    Not yet. Local Services Ads still produce roughly half of signed cases in legal accounts with meaningful ad spend, and AI Overviews appear on only 7.9% of local searches. Fund AI visibility out of your existing SEO and content work instead.

    If you want the tactical version of the AI visibility work, our guidebook How to Get Your Law Firm Cited in AI walks through the schema, content structure, and citation sources that get firms named inside AI answers. And if the foundation underneath it needs work first, that’s what our law firm SEO team does every day.

    The firms that win the next two years won’t be the ones who guessed the right chatbot. They’ll be the ones who were already easy to find, easy to verify, and easy to reach.

    Share on Social Networks