Original research · September 2026 · First Mind·AI
The Answer Share
Which PI firm gets recommended by AI in your city?
More and more people are turning to AI chatbots when they need an attorney. I wanted to find out who owned the answers those bots were giving back. So I asked three of the most popular AIs, ChatGPT, Gemini and Perplexity, the 10 questions a potential client would ask, and I asked those same questions across 15 of the biggest legal advertising markets in the US. Some of the findings were surprising. Some weren’t. The main takeaway is this: if your firm isn’t paying attention to whether AI is naming you, you are losing market share to a firm that is.
10 household names scored 0% of AI answers in their own backyards.
The billboard on the drive home. The jingle you can still sing. The name a local produces without stopping to think. We put the ten questions their own clients ask to the assistants, 30 answers per market, and not one of these firms came back.
Why this matters now
For twenty years, a person who got hurt opened Google, scanned a page of ten blue links, and picked one. That behavior is coming apart fast, and it is coming apart specifically in the moment that matters to a personal injury firm: choosing who to call.
Source: iLawyer Marketing, 2026 consumer study, 1,110 U.S. adults aged 18 to 65, asked which online sources they would use to research and find an attorney. These are stated intentions rather than observed sessions, and iLawyer Marketing is a legal marketing agency reporting on its own category. We cite it because it is the only recurring consumer-side series on this question, and because its direction matches what we measure directly: the numbers in this study come from querying the assistants themselves.
Not a new ranking. A different system.
The instinct is to treat this as one more channel to rank in. It is not. The mechanics differ at almost every step, which is why firms that have won at search for a decade can hold a 0% answer share.
A search engine
Ten results · the user chooses
- ›Returns a list. Being on page one is enough to be considered.
- ›Ranks pages against a keyword, recomputed per query.
- ›Being the best source earns the click.
- ›You can buy your way onto the page.
An answer engine
One answer · two to five names
- ›Returns a shortlist. There is no page two to be on.
- ›Names entities it can verify, drawing on what it already knows about your firm.
- ›Being the source does not mean being named.
- ›There is no placement to buy.
The third line is the one that costs firms the most, because it is invisible from the inside. We have watched an assistant read a firm's own page, use it to answer the question, and then describe the author as “one local car-accident law firm” before sending the reader to a bar association. The firm supplied the answer and received none of the credit. In search that is a win. In the answer layer it is a loss, and nothing in a rankings report will show it to you.
The visitors are also worth more. Semrush, analyzing more than 500 high-value commercial topics, found that people arriving from AI search convert at roughly 4.4 times the rate of ordinary organic traffic. They arrive having already been told who to consider, which lands them further down the decision than a searcher still building a shortlist. And the citation pool is narrower than search rankings ever were: page one offered ten slots with more pages behind it, while an answer names two to five firms and stops. The rest of this page shows who is getting into that answer, market by market, and who is not.
Conversion figure: Semrush, published June 2025, from an analysis of over 500 high-value commercial topics. It is a cross-industry benchmark rather than a legal-sector measurement, and we cite it as an indication of direction, not a number any individual firm should forecast against.
The blackout
A 0% answer share means the firm was not named in any of the 30 answers recorded in that market on the date sampled. It measures AI answer frequency only. It is not a statement about any firm’s results, reputation, or quality of representation.
What the split actually tracks
The cleanest way to see what actually separates a named firm from an unnamed one is to put two of them side by side inside a single market, answering the same ten questions on the same day. Both are long-established Jacksonville injury practices. Only one of them is legible to the assistants.
Pajcic & Pajcic
Jacksonville · 40% answer share
- +Named in 12 of 30 answers, by two of the three assistants.
- +Returned unprompted for broad questions, not just when asked about them by name.
- +Perplexity misspelled their domain and still resolved it to the firm. The assistant knows the practice, not just a URL.
Terrell Hogan
Jacksonville · 0% answer share
- –Named in 0 of 30 answers, by any of the three assistants.
- –Equally established locally, and equally invisible to every engine.
- –Nothing for the assistants to retrieve, so nothing comes back.
Same city, same clients, same ten questions. Twelve mentions to zero. The difference is not spend and it is not tenure, because both firms have both. It is that one of them has left enough verifiable trace across the open web for a model to retrieve and repeat, and the other has not. A TV buy purchases attention from people. It purchases nothing from a model that never saw the ad. Worth noting even the winner here is not winning everywhere: ChatGPT named Pajcic zero times. Answer share is won engine by engine, and almost nobody is covering all three.
01 · St. Louis, MO
New this editionThe highest answer share recorded anywhere in this study, and it belongs to a television advertiser.
Brown & Crouppen took a 63.3% answer share, nearly double the next firm and the highest figure in the study. They are a heavy television advertiser. This is the clearest evidence here that ad spend and answer share are not opposed. They are simply unconnected, and a firm that gets the underlying signals right wins both.
02 · Philadelphia, PA
New this editionNo firm owns this market, and the two names Philadelphians know best are both absent from it.
Lundy Law, the firm behind “Remember this name,” holds a 0% answer share. So does Rand Spear, “The Accident Lawyer.” Eric Shore managed 2 of 30. Two of the most heavily advertised injury brands in Pennsylvania are invisible to the assistants their clients now ask, and nobody in the market clears 25%. Philadelphia is genuinely unclaimed.
03 · Miami, FL
New this editionThe most fragmented major market measured: a three-way tie at the top and a famous name at zero.
Rubenstein Law, on Miami television and radio for over 35 years, holds a 0% answer share. Anidjar & Levine managed 1 of 30. Even Morgan & Morgan, which takes 53% in Orlando and 60% in Tampa, collapses to 2 of 30 here. Statewide brand strength does not carry into a market where the assistants have no local signals to attach it to.
04 · Houston, TX
New this editionTrial firms lead. The most recognizable injury brand in Texas sits near the bottom.
Jim Adler, “The Texas Hammer,” took a 6.7% answer share in Houston, the identical figure he posted in Dallas. Two editions, two of the largest legal ad markets in America, and one of the most recognizable legal brands in the country lands at 2 of 30 in both. Terry Bryant, a long-running Houston TV advertiser, managed 4.
05 · Chicago, IL
New this editionA genuinely competitive market, and the one place where advertising and answer share partly overlap.
Chicago is the healthiest of the new markets: a TV advertiser leads and the elite trial firms sit close behind. The outlier is Rosenfeld Injury Lawyers, a substantial local advertiser, at 1 of 30. Even in a functioning market a firm can be almost entirely missing from the answer set.
06 · Jacksonville, FL
New this editionThe healthiest market in the study, and the sharpest same-market split between two established firms.
Jacksonville is where reputation and answer share line up: the firms people name are the firms the assistants name. The exception is decisive. Terrell Hogan, one of the city’s longest-established injury practices, holds a 0% answer share while Pajcic & Pajcic holds 40%. Same city, same clients, same ten questions. Twelve mentions to zero.
07 · West Palm Beach, FL
New this editionThe flattest market measured. Nobody exceeds 20%, and the leading advertiser is effectively absent.
Gordon & Partners, whose “For The Injured” brand is among the most advertised on the Treasure Coast, appeared once in 30 answers, and only under their firm-name domain rather than the brand they spend money on. The same firm takes a 23.3% answer share in Miami, a market they do not anchor: they are seven times more visible outside their home market than inside it. Steinger, Greene & Feiner and Lesser, Lesser, Landy & Smith each managed 2. No firm here has established itself, which makes this the most winnable market measured.
08 · Tampa, FL
July market, re-scoredMorgan & Morgan’s strongest market anywhere.
Winters & Yonker and Florin|Roebig, both sustained Tampa Bay advertisers, hold a 0% answer share. Morgan & Morgan takes 60% of the market on its own.
09 · Orlando, FL
July market, re-scoredThe #1 legal advertising market in America, and two brands take most of the oxygen.
Michael T. Gibson, the “Auto Justice Attorney,” and Rue & Ziffra both hold a 0% answer share in the country’s largest legal ad market.
10 · San Antonio, TX
July market, re-scoredA trial boutique still leads the second-highest spot-TV spender in America.
Wayne Wright, a San Antonio television fixture for decades, holds a 0% answer share. Thomas J. Henry, at roughly $21.5M a year in spot TV, still trails a trial boutique 40% to 50%.
11 · Dallas, TX
July market, re-scoredThe widest gap in the study between advertising presence and answer share.
Five heavily advertised Dallas firms hold a 0% answer share: Thompson Law (“Ryan The Lion”), Witherite Law Group (1-800-CarWreck), The Barber Law Firm, Loncar Lyon Jenkins, and Ben Abbott & Associates. Jim Adler managed 2 of 30.
12 · Atlanta, GA
July market, re-scoredThe rare market where a legacy TV brand stays near the top.
Kenneth S. Nugent (“One call, that’s all”), The Champion Firm and Witherite Law Group each hold a 0% answer share. Montlick, on Atlanta billboards for four decades, managed 2 of 30.
13 · Phoenix, AZ
July market, re-scoredA local boutique edges out one of the biggest advertisers in the Southwest.
Sweet James and Torgenson Law, both significant Arizona advertisers, hold a 0% answer share.
14 · Las Vegas, NV
July market, re-scoredThe most fragmented of the original markets, and the best-known names sit near the bottom.
Richard Harris Law Firm and Eglet Adams, two of the best-known names in Nevada injury law, each took a 6.7% answer share. Dimopoulos Injury Law and THE702FIRM managed 1 of 30 apiece.
15 · Kansas City, MO
July market, re-scoredNo dominant presence at all, and the sharpest example anywhere of a firm losing its identity between two brand names.
Five established Kansas City firms hold a 0% answer share: Wendt Law Firm, Krause & Kinsman, Devkota Law Firm, Peterson & Associates and Bertram & Graf. The most instructive result belongs to DM Injury Law, formerly DiPasquale Moore, one of the largest advertisers in the metro. They appeared once in 30 answers, and that single mention surfaced under dipasqualemoore.com, the name they have moved away from. Their current domain, dmlawusa.com, was named zero times. The assistants retained the old entity and never connected it to the new one, which is the clearest illustration here of what a split identity costs. Brown & Crouppen, which takes 63% of neighboring St. Louis, drops to 20% here.
The next market run
Ready to be the answer AI gives in your city?
Every market on this page has an opening in it. Thirty minutes is enough to see where your firm stands today, which competitors are taking the answers you want, and what it would take to change that before the next edition runs.
Book a 30-minute call with AdamMethodology
Seven markets were collected on September 1, 2026; the original seven and Kansas City were collected on July 12 and July 28, 2026 and re-scored from their stored responses. All runs used US geo. The same 10 discovery-stage questions go to three AI platforms, ChatGPT, Gemini and Perplexity, for 30 answers per market, 450 in total. A firm’s answer share is the percentage of its market’s 30 answers that name the firm. Brand variants and misspelled domains are merged to the firm entity; directories, bar associations, insurers, hospitals and government resources are excluded.
The 10 questions
- “best personal injury lawyer in {city}”
- “best car accident lawyer in {city}”
- “who should I call after a car accident in {city}”
- “should I hire a lawyer after a car crash in {state}”
- “how do I find a good personal injury attorney in {city}”
- “top rated accident and injury law firms in {city}”
- “best truck accident lawyer in {city}”
- “lawyer to help with an insurance claim after an accident in {city}”
- “I was injured in an accident in {city}, which law firm should I use”
- “which law firm wins the biggest injury settlements in {city}”
What changed since the July edition. The July study used four platforms, including Grok. Grok was removed from our data source between editions and can no longer be queried, which means that study is no longer reproducible exactly as published, by us or by anyone checking our work. Rather than quietly swap in a different engine and present the result as continuous, we re-scored every market, including the original seven, on the three platforms that remain, using the same stored responses. Every figure on this page is comparable to every other figure on this page.
AI answers vary with time, location, and phrasing; this is a same-method snapshot, not a guarantee of what any individual will see. Answer share measures frequency of mention in AI-generated answers and implies nothing about the quality of any firm’s legal services. Firms named at or near zero were domain-verified at the time of publication. A small number of AI-surfaced domains we could not attribute to a verified firm are excluded from the tables rather than guessed at. All trademarks belong to their owners; no firm listed is affiliated with First Mind AI unless stated. Corrections or data requests, including re-testing of any firm named here: adam@firstmindai.com.
What the assistants actually read
AI assistants recommend firms they can verify: a consistent identity everywhere they look, third-party authority that vouches for you, and published answers to the exact questions injured people ask. Brand awareness does not transfer, which is why a boutique with clean entity signals outranks a $20M TV budget, and why a television advertiser that gets the signals right wins both layers at once. The opportunity is in how open this still is: in Miami, Philadelphia and West Palm Beach, no firm holds even a quarter of the answers.
Frequently asked
What is answer share?
Answer share is the percentage of AI answers in a market that name a given firm. It is the AI-era counterpart to share of voice: instead of measuring how often your advertising reaches a person, it measures how often an assistant names you when that person asks who to call. A firm named in 12 of a market’s 30 answers has a 40% answer share.
What is The Answer Share study?
A recurring study by First Mind AI measuring which personal injury law firms AI assistants actually recommend, metro by metro. The September 2026 edition covers 15 U.S. legal advertising markets and 450 AI answers to the 10 questions injured people ask most, submitted to ChatGPT, Gemini and Perplexity.
Which personal injury law firms have the highest answer share?
Brown & Crouppen in St. Louis leads at 63.3%, named in 19 of 30 answers. Morgan & Morgan leads Tampa (60%) and Orlando (53.3%). Elsewhere the leaders are trial firms rather than the biggest advertisers: the Law Offices of Frank L. Branson in Dallas and the Law Offices of Pat Maloney in San Antonio (both 50%), Pajcic & Pajcic in Jacksonville (40%), and Arnold & Itkin in Houston (36.7%).
Which well-known law firms are invisible to AI assistants?
Ten firms with substantial advertising presence held a 0% answer share in their market: Lundy Law and Rand Spear in Philadelphia, Rubenstein Law in Miami, Terrell Hogan in Jacksonville, Kenneth S. Nugent in Atlanta, Winters & Yonker in Tampa, Thompson Law and Witherite Law Group in Dallas, Wayne Wright in San Antonio, and Sweet James in Phoenix. Jim Adler, one of the most recognized legal brands in Texas, took a 6.7% answer share in both Dallas and Houston.
Does TV and billboard advertising improve a law firm’s answer share?
Our data shows no relationship in either direction. Many of the heaviest advertisers hold a 0% answer share, while Brown & Crouppen, also a heavy television advertiser, posts the highest figure in the study at 63.3%. Advertising neither causes nor prevents AI visibility. The two systems read entirely different signals.
How is answer share calculated?
Each market is a same-method snapshot: 10 standardized client-discovery questions submitted to 3 AI platforms, for 30 answers per market. A firm’s answer share is the percentage of those 30 answers that name the firm. Brand variants and misspelled domains are merged to one entity, and directories, bar associations, insurers, hospitals and government resources are excluded.
Why does this edition use 3 platforms when the July edition used 4?
The July edition included Grok. Grok was removed from our data source between editions and can no longer be queried, so that study is not reproducible exactly as published, by us or by anyone checking our work. Rather than quietly substitute a different engine, we re-scored every market, including the original seven, on the three platforms that remain, using the same stored responses. Every figure here is directly comparable to every other figure here.
How can a personal injury law firm improve its answer share?
AI assistants recommend firms they can verify: a consistent identity across the web, third-party authority such as news coverage and citations, and published answers to the exact questions injured people ask. Brand awareness does not transfer. First Mind AI offers a free AI visibility audit that shows a firm its current answer share and what is holding it back.
See how AI sees your business.
Find out whether ChatGPT, Gemini, and Perplexity recommend you, and exactly where you are losing visibility. Enter your domain and email to get your instant analysis.
See exactly what a client sees.
Every First Mind client gets a live dashboard: where their AI visibility stands, why it moved, and what ships next. This is that dashboard, running on a sample firm, with the same layout, controls, and reasoning, and invented numbers.
- Where the firm stands this month, on the real questions buyers ask
- Which competitors AI names instead, and by how much
- The approval queue: every draft, one click to approve or send back
- The monthly report, rebuilt from live measurement rather than typed up