Because AI engines do not recommend the firm people have heard of. They recommend the firm they can confirm. When someone asks ChatGPT or Google's AI Mode for a lawyer, the engine pulls directories, profiles, reviews and firm pages for that practice area and city, then names the firms those sources agree on. A billboard is not one of those sources. A complete Avvo profile is.
My position, for a partner at a small or mid-size firm who keeps losing AI answers to a name on every bus bench: fame is not the thing to match. Evidence is, and evidence is something a smaller firm can build.
How does an AI engine decide which firm to name?
It searches, reads and names what the sources support. Google's documentation says AI Overviews and AI Mode may use "query fan-out", running several related searches across subtopics and data sources before writing the answer and choosing the links to show. So "best divorce lawyer near me" becomes a handful of lookups: who practices family law here, who has reviews, who appears in the directories, who has a page about the specific matter.
What comes back is mostly third-party material about firms. In BrightLocal's 2026 study of local AI answers, the lawyer category drew 13,354 citations across Google AI Overviews, AI Mode and ChatGPT, and the most cited sources were google.com (42%, which in that study is largely Business Profiles), Yelp (7.9%), Justia (3.2%), Avvo (2.5%) and Lawyers.com (1.8%). BrightLocal sells listing management and the lawyer sample is small, so treat the shares as a map of where to look, not a formula.
None of those sources measures how many people have seen your TV spot. They record what your firm does, where, for whom, and what clients said. An engine that finds the same facts in five places has something it can repeat with confidence. An engine that finds a famous name with thin, conflicting profiles has less to go on.
What the law firm studies show about who gets named
The clearest sign that fame is not deciding these answers is how widely the recommendations spread. Custom Legal Marketing ran 3,200 prompts such as "recommend a personal injury lawyer in" a given city through ChatGPT, Claude and Google AI Mode in July 2026, covering 100 metros and 32 practice areas. The answers named 19,692 different law firms. Only 578 of them (2.9%) appeared on all three platforms, the three agreed on the first firm named in 0.9% of prompts (24 of 2,566), and 15,308 firms appeared in only one city. The study found no dominant national name.
The same study shows why: each engine reads a different shelf. Avvo appeared in 87.1% of Claude's answers, 29.2% of ChatGPT's and 5.4% of Google AI Mode's. Super Lawyers appeared in 62.8% of AI Mode's answers and none of Claude's. Justia appeared in 60.2% of AI Mode's answers and Google Business listings in 36.8%. A firm that is well documented on the shelf a given engine reads gets named by that engine, whatever its size.
A second study points the same way. Forward Push, a law firm marketing agency, logged 38,173 answers from ChatGPT, Perplexity, Gemini and Claude between late April and early June 2026 across 25 practice areas and 15 cities. It reported that ChatGPT and Perplexity overlapped on only about 6% of firms for the same question, and that 96% of Perplexity's answers cited at least one legal directory while 68% cited the firm's own website. Both studies come from firms that sell marketing to lawyers, and both describe a snapshot. Read them as direction, not as law.
Where fame helps, and where it stops
Being well known does matter, but only in the form the engines can read. Ahrefs studied 75,000 brands in 2025 and found that branded web mentions had the strongest correlation with how often a brand appeared in Google's AI Overviews, at 0.664. Branded search volume came in at 0.392. Spending on branded search ads came in at 0.215, close to the bottom of the list. Ahrefs says plainly that correlation is not causation, and the study covers Google's AI Overviews, not ChatGPT, and brands in general, not law firms.
The pattern is still useful. Fame that turns into people writing about you, linking to you and reviewing you shows up in the answer. Fame that lives only on a highway sign or in a 30-second spot does not, because there is nothing for the engine to read. Google's own local ranking help page puts it in the same terms: prominence means how well-known a business is, and it is "based on info like how many websites link to your business and how many reviews you have". The same page says there is no way to pay for a better local ranking.
A worked example: two firms, one question
For example, suppose a large personal injury firm runs billboards across a metro and a five-attorney firm handles mostly trucking cases in two counties. A prospective client asks an AI engine, "Who handles truck accident cases in" the smaller county.
Suppose the large firm's Avvo and Justia profiles list practice areas in general terms, its Google profile is tied to a downtown office 40 miles away, and its site has one page that covers every kind of vehicle accident. Now suppose the smaller firm has a Google Business Profile at a real office in that county, reviews on Google and Yelp that mention trucking cases, a Justia profile and state bar listing with the same attorneys, address and practice areas, and a page that states plainly that it represents injured people in commercial truck crashes in those two counties, on contingency.
In this illustration, every lookup the engine runs for that question finds the smaller firm, and finds the same facts each time. The large firm is better known and still gives the engine less to confirm for this specific question. This is a hypothetical, not a client result, and no one can promise how a given engine will answer. It shows where a small firm's advantage sits: in narrow, specific questions where evidence beats reach.
"But the big firms still get named more often"
This is the strongest objection, and parts of it are true. In Firesign's informal July 2025 test of one query, "commercial litigation firms in Kansas City," three of the five AI tools it tried favored large firms: two versions of ChatGPT and Gemini put mostly large firms at the top. The author tied this partly to the sources those tools cited, such as firm Wikipedia pages and Chambers rankings, which favor big firms by design. Firesign's 2026 analysis for a boutique client found ChatGPT citing law firm Wikipedia pages in 35.6% of queries. Forward Push found the top three firms in a city and practice area took about a third of mentions. And the Ahrefs data says brands with more web mentions get more AI visibility. Big firms have more of everything.
Here is the answer. First, look at what drove the big-firm tilt: Wikipedia, Chambers and national rankings. Those are evidence, not fame. A 1,000-lawyer firm has a Wikipedia page because journalists and rankings wrote about it, and an engine reads that record. It does not reward the billboard. Second, the same Firesign test found the opposite result elsewhere: Microsoft Copilot named only firms of 1 to 17 lawyers, and Google's AI Overview put small firms in its first 13 places. Third, that test was one commercial litigation query in a market with eight AmLaw 100 firms. Consumer matters in a single county are a different contest, which is what the 15,308 single-city firms in Custom Legal's study reflect.
So the big firms do win the questions where national records decide the answer. A smaller firm should not fight there first. It should own the questions where the record is local and specific: the practice area, the matter type, the county, the fee basis. That is where the evidence is cheap to build and the large firm's advantage is thin.
What should a smaller firm do first?
Start with the evidence the engines already read, in this order:
- Make sure the engines can read your site. Check robots.txt and your firewall or CDN settings for Googlebot, OAI-SearchBot and PerplexityBot. A blocked crawler makes everything else irrelevant for that engine.
- Make every profile say the same thing. Google Business Profile, Avvo, Justia, Martindale, Lawyers.com, FindLaw, Super Lawyers if you have it, and your state bar listing: the same attorneys, office address, phone and practice areas. Remove lawyers who have left.
- Write one page per matter you want to be named for. State who you represent, in which matters, in which counties and on what fee basis, in plain sentences an engine can lift. Keep any results claim next to the context your state's version of Rule 7.1 requires, and do not call anyone a specialist unless your state allows it.
- Ask for reviews on more than one site, within your state's rules: nothing of value in exchange, and no scripted wording.
- Earn mentions that describe the work. Bar association talks, local press, a quoted comment on a case in the news. These become the web mentions that the fame data actually measures.
Our post on how law firms get recommended by ChatGPT covers each source in more detail, and AEO for law firms explains how to measure progress without a ranking report.
How to check where you stand
Pick the five questions your intake team hears most, add your county, and ask each one in ChatGPT, Claude, Perplexity and Google's AI Mode. Write down which firms are named and which sources are linked. Then open the profiles of the firm that keeps beating you and compare them with yours, field by field. In most cases the gap you find is something you can fix, not a reputation you have to buy.
If you want that comparison done for you, our free AI visibility audit runs your practice-area questions across five engines and shows which firms are named and which sources the answers rely on. More on how we work with firms is on the law firms page.
Frequently asked questions
Can a solo or small firm really be named over a large firm?
Why does ChatGPT name a competitor and not us?
Should we stop billboard and TV advertising?
Which directory matters most for AI answers about lawyers?
Sources
- Do ChatGPT, Claude, and Google AI Mode Recommend the Same Law Firms? · Custom Legal Marketing
- Google Ranks, AI Recommends: New Study of 38,000 AI Searches Reveals Which Law Firms Actually Get Named · Forward Push Law Firm Marketing (press release via WBOC)
- Law Firms and AI Recommendations: Do LLMs Favor Big Law? · Firesign via JD Supra
- AI and Law Firm Recommendations: What are LLMs Actually Crawling? · Firesign via JD Supra
- An Analysis of AI Overview Brand Visibility Factors (75K Brands Studied) · Ahrefs
- Tips to improve your local ranking on Google · Google Business Profile Help
- AI Features and Your Website · Google Search Central
- Top sources and directories for local AI search: What 1.9 million citations across 116,670 domains tells us · BrightLocal