How a law firm gets named in Google AI Overviews and AI Mode
Written by
Odin Lund
Category
Law firms
Date
Share this article
What do Google AI Overviews read before they name a lawyer?
Being named in Google AI Overviews means that when a person searches Google for a car accident lawyer and the results page opens with an AI-written summary above the ordinary links, that summary includes the firm by name, usually with a short line on it and a link in the citation panel beside the text. In AI Mode, Google's conversational search tab, being named means the same thing inside a longer written answer that a person can keep questioning.
Both are built on Google's own index. For a question like this Google runs the search, or several related searches at once in AI Mode, pulls a set of pages and videos it already ranks for those searches, reads them, and writes a summary from what they say. The three steps are the same as on any engine. Retrieval decides which pages are read. Selection decides which firms on those pages get named. Synthesis decides what is said about each one. The difference on AI Overviews is in step one. The pages Google reads for these questions are not the pages ChatGPT reads, and our data makes that plain.
Our tracker ran 111 New York car accident questions every day between 18 September and 17 October 2026 for William Mattar Law Offices, the firm we manage in that category, against 13 competing firms on ChatGPT, Gemini and Google AI Overviews. Of the 9,049 answers, 2,389 came from AI Overviews. The count is lower than on the other two engines in part because Google does not show an Overview for every query every day, and we only record an answer when one appears. For each Overview we recorded every source it retrieved and cited. Ranked by the share of answers in which each one was retrieved, the diet looks like this. One note on the words. Retrieved means the engine fetched the page while writing the answer. Cited means it attached that page to a line of the answer. The two are not the same, and we record both. The shares below are retrieval shares, because retrieval is the step a firm has to clear first.
YouTube, 23% of answers. The single most retrieved source on this engine. Short explainer videos on what to do after a crash, how no-fault works and whether to hire a lawyer, most of them posted by law firms.
A competitor firm's site, 19%. One firm's own pages, retrieved in nearly one answer in five. Competitor firm sites as a group were retrieved in 10% to 19% of answers each.
The firm's own site, 14%. William Mattar's practice pages and process answers, third on this engine even though they lead on the other two.
Government sources, 7% combined. The state and court pages that ChatGPT reads first appear in fewer than one Overview in ten.
Facebook, 6%. Firm pages and posts, and the reviews and comments attached to them.
Justia, 6%, and Avvo, 5%. The directories that drive ChatGPT are present here, but far down the list.
Reddit, 5%. Threads in which people describe their own accident and ask who to call, with firm names in the replies.
Instagram, 5%. Firm profiles and posts, read for the firm's name, its city and what it says it does.
What this means is that AI Overviews are built from video and social proof, not from the state and not from the directories. The engine that sits on top of Google itself is the one least influenced by the website and the directories, and the one most influenced by what a firm has published on YouTube, Facebook, Reddit and Instagram.
Which questions put a firm in the answer?
The 111 questions we run fall into five groups, published in full in our companion post. Fifty-one ask the engine to find or recommend a lawyer. Twenty-six ask how a claim works in New York. Ten ask whether hiring a lawyer is worth it and what it costs. Eight ask the engine to compare firms. Sixteen name one firm and ask about it. On Google the groups behave differently from the way they behave on ChatGPT, because the sources are different.
The finding questions are still the ones that return named firms, and most of them carry a city. When an Overview appears for one of these, the names in it come from the firm sites and videos Google retrieved, not from a directory list, so the firms named tend to be the ones with a video or a strong page for that city and case type. In the most recent 30 days William Mattar was named in 9.6% of AI Overview answers against 22.3% on ChatGPT, for the same questions, which is the gap the source diet predicts.
The process questions are the surprise on this engine. ChatGPT answers them from the state, but government sources appear in only 7% of Overviews. Here the explanation of no-fault, serious injury and filing deadlines is assembled from firm pages and firm videos, which means a process answer on Google is also a naming opportunity. The firm whose page or video explains the rule is the firm cited next to it. The whether-to-hire questions draw on the same firm pages and on the community layer, Reddit threads and Facebook comments where people describe what happened when they did or did not hire someone. The comparison questions draw on firm sites and the reviews attached to social profiles, and the answers tend to list firms rather than rank them. The firm-name questions return a summary of that firm built from its site, its video channel and its social profiles together, so a stale Facebook page or an old video can shape what Google says about a firm today.
In the customer's words, the questions sound like this.
Best car accident lawyer in Buffalo for a serious injury case
car accident lawyer near me no fee unless we win
What are my options after a hit and run car accident in New York?
the adjuster offered me a settlement for my neck injury, should I take it or get a lawyer
who did you use for a car accident on Long Island
William Mattar reviews
Two of those six come from our published list. The others are written the way people type into a search box rather than the way they speak to an assistant, which is the main difference between Google's questions and ChatGPT's. Shorter, lower case, often with a number or a city and nothing else. In AI Mode the phrasing drifts back toward full sentences and follow-ups, and the same sources carry the answer.
What a firm changes to get named
The work for Google follows the Overview diet, which is the opposite of the ChatGPT diet. Each item below is a change to a source Google retrieves, in the order we do it.
Build the YouTube layer. YouTube was retrieved in 23% of Overviews, more than any other source. We publish short explainers on the exact questions in the set, one video per question, with the firm name and the city in the title and the description, and an attorney answering in the first minute. The transcript should say what the matching page on the site says, so the two sources agree.
Make Facebook and Instagram state the same facts as the site. Name, office cities, practice areas, fee line and consultation terms, identical to the practice pages. Facebook also carries reviews that Google reads, so the review flow that most firms point at Google should point at Facebook as well.
Be present where people ask each other. Reddit was retrieved in 5% of Overviews, and the names in those threads are the names people give each other. We do not post as clients and we do not ask firms to. What we do is make sure the firm's published answers are the pages people link to when they reply, and that the firm's own account answers the general questions where a community's rules allow it.
Let the practice pages carry the process answers. Because government sources appear in only 7% of Overviews, firm pages do the explaining on this engine. A page that answers how no-fault works in plain language, with the statute and the numbers, is retrieved directly, and the firm is cited with it.
Study the competitor whose site leads. One competitor's site was retrieved in 19% of Overviews. Our tracker shows which of its pages Google pulls and for which questions. We publish better answers to the same questions, in text and in video, rather than guessing.
Measure AI Overviews on their own line. For the same firm and the same questions, share of voice was 54.6% on ChatGPT and 20.8% on AI Overviews. A blended number averages those into something that describes neither engine.
Benchmark where you stand before any of this. Take twenty of the questions your clients ask, phrased the way they would type them into Google, and run them from a logged-out browser. For each one, note whether an Overview appears at all, which firms it names, and which sources sit in the citation panel. Run the same twenty in the AI Mode tab and note the same three things. After a week you will know whether you are losing on video, on social, or on the page itself. If you would rather we did it, request a free preview audit at audit.sourcenode.co and we will send back what Google says about your firm today and where it comes from.
This is what the work looked like for William Mattar Law Offices, across all three engines. When we started, the firm held 4.6% share of voice and was named in 2.4% of relevant answers, fourth of 14 firms. In month one we rewrote the main practice pages so each answered one customer question with the same set of facts, and share of voice reached 29.5%. In month two we published 31 new pages against the questions where no firm was yet named and worked the directories and publications the engines cited most, and share of voice reached 43.8%. Month three was spent holding the position, and share of voice closed at 47.2% with visibility at 23.3%, first of 14. On AI Overviews specifically, in the most recent 30 days the firm holds 20.8% share of voice, is named in 9.6% of answers, and sits at an average position of 2.3 when named. It is first on this engine too, but by a thin margin over the competitor whose site Google retrieves most, and that is the open flank the video and social work is now closing. The full method and its limits are on our results page, and our law firm page describes how an engagement runs week by week.
Common questions
Is AI Mode the same thing as an AI Overview? Same systems, different surface. An Overview is a summary at the top of an ordinary results page. AI Mode is a separate tab that answers in full and takes follow-up questions, and Google has described it as running several searches at once behind each answer. Our tracker records Overviews. Because both draw on the same index, the source work described above serves both, and we check AI Mode by hand alongside the daily run.
Do we need to rank first on Google to be in the Overview? Ranking is not the test. The Overview is written from a set of pages and videos Google pulls for the question, and in our data the most retrieved source was YouTube, which is not a page ranking at all. A firm with no video and no social footprint is competing for the Overview with one hand, whatever its organic rankings, and a firm with both can be named for questions where its site does not rank.
Why is our share so much lower on AI Overviews than on ChatGPT? Because the two engines read different sources. The work most firms have done, on the site and on the directories, is exactly what ChatGPT and Gemini read, so those engines reward it. AI Overviews reach for YouTube, Facebook, Reddit and Instagram, where most firms have a smaller footprint than the one or two competitors who have invested there. The gap is not a ranking problem. It is a missing source group.
Can you guarantee that Google AI Overviews recommend us? No, and anyone who does is selling something else. We do not control the engine, and Google does not show an Overview for every search. What we commit to is the measurement, the work on the sources Google retrieves, and a published account of what moved and what did not. One documented engagement is one engagement, and categories differ in how contested the answers are.
Do the videos need to be professionally produced? No. What we work on is the title, the description, and what the attorney says in the first minute, because those are what carry the firm's name, its city and the answer to the question. A clear answer recorded on a phone in the office does the job. What matters is that each video answers one question from the set and matches the page on the site that answers the same one.
Sources. SourceNode, Who do AI engines trust when someone asks for a car accident lawyer in New York? 9,049 answers, three source diets, October 18, 2026. SourceNode, 111 questions people ask AI before they call a car accident lawyer in New York, October 18, 2026. SourceNode, Measured results, sourcenode.co/results, data to 17 September 2026. Our tracker, 14 New York motor vehicle accident firms, 30 days to mid-October 2026. Google, Expanding AI Overviews and introducing AI Mode, March 5, 2025.
Share Blog
