Writer Login Client Login Order Online

Personal Injury AEO: How to Write a Fee Page Answer Engines Will Quote

Personal Injury AEO: How to Write a Fee Page Answer Engines Will Quote

Posted on by David B. Arato

Every answer engine already knows roughly what a personal injury lawyer charges. Ask any of them and you get the same shape of answer: a percentage of what the client recovers, stepping up once a lawsuit is filed.

So a firm cannot win the fee question by explaining contingency fees better than the firm down the road. What separates the pages that get quoted from the pages that do not is whether a specific figure appears at all, how early it appears, and whether it says what it actually covers. Personal injury AEO across answer engines turns on precision about money more than on anything else on the page.

Q Short Answer

A personal injury fee page should state a specific percentage near the top, give the pre-suit and post-filing rates separately, say what happens if the case is lost, and say whether expenses come off before or after the fee.

Of ten quoted firm pages, one answered that last question plainly.

Put the Percentage on the Page

A personal injury fee page needs a number on it, and the absence of one is the clearest difference between the pages quoted on fees and the pages that were not.

Five of the ten stated a specific contingency percentage. All five were pages an engine had quoted on the fee question. The other five, quoted on different questions, stated no rate of their own anywhere on the page an engine used.

A range counts. A page giving its own percentage, or its own range tied to the stage of the case, is stating a number. A page saying the fee is a percentage of what the firm recovers is not.

A percentage published on a fee page falls under ABA Model Rule 7.1 as a communication about the firm’s services, which is why the sentence carrying it should name the stage it applies to and what comes out of it. Precision is the compliant answer here. Silence is not.

Publish Both Rates, Not Just One

Contingency fees usually step up once a lawsuit is filed, and a page that gives one number is describing half of what a client will pay.

Three of the ten pages published the split, a lower rate if the matter resolves before a lawsuit and a higher one once litigation starts. All three were pages quoted on the fee question. None of the five quoted on other questions published it.

The split also gives a reader something a single figure cannot, which is a reason the number moves. A passage that explains itself is a passage that survives being lifted out of the page.

Say Whether Expenses Come Off Before or After the Fee

Whether case expenses come off before or after the percentage is the question almost nobody answers, and the order changes how much money a client actually receives.

Case expenses are separate from the attorney’s fee. They are billed on top of it and reimbursed out of the recovery:

  • filing fees
  • medical records
  • expert witnesses
  • depositions

Subtracting them before the percentage is calculated produces one number for the client. Subtracting them after produces another. Someone comparing two firms has no way to tell which they are being offered unless the page says.

Of the ten pages, one stated the order plainly and showed the arithmetic. One other stated it in two places and contradicted itself. The remaining eight said only that expenses are separate.

A page that answers this is answering something a reader is actively trying to find out and getting nowhere with, which is rare in a practice area where every topic has been written a thousand times.

Say What Happens If the Case Is Lost

State what a client owes when there is no recovery, and say separately whether that covers expenses as well as fees.

Five of the ten pages addressed it. The strongest were plain sentences: no attorney fee is owed if the matter does not succeed, and here is whether the client is responsible for costs already advanced. Several pages gestured at it with a slogan instead, which is weaker to read and weaker to quote.

A Fee Page Without a Figure Is Not a Fee Page

A fee-focused web address does not get a page quoted on fees by itself. Two of the ten sat at a dedicated fee URL, carried no percentage anywhere, and were not quoted on the fee question.

Placement did more than the slug. Two of the quoted pages put a figure in the first screen, above the explanation, and no page in the other five had a number in the first screen at all. A reader who has to scroll past four paragraphs to reach the point is being asked to work for it, and so is anything trying to lift a short answer out.

Markup does not do the work either. According to Google’s optimization guidance for generative AI features, no special schema.org markup is needed for generative AI search, so a fee page with complete structured data and no percentage still gives an engine nothing to lift. Keep the schema for classic search results, and put the figure on the page for everything else.

Frequently Asked Questions

Should a personal injury firm publish its contingency percentage?

A personal injury firm should publish its contingency percentage. All five firm pages quoted on the fee question in this set carried a specific figure, and none of the five quoted on other questions stated a rate of its own. Because the figure on the page is a communication about the firm’s services, name the stage it applies to and say what comes out of it.

What is the most common gap on a personal injury fee page?

The most common gap on a personal injury fee page is whether case expenses are deducted before or after the attorney’s percentage is calculated. One of ten quoted firm pages answered it plainly, and the two orders produce different amounts of money for the client.

Does putting fees in the web address help a personal injury page get quoted?

Putting fees in the web address does not help on its own. Two of ten quoted firm pages sat at a dedicated fee address, carried no percentage, and were not quoted on fee questions. The figure on the page did the work rather than the slug.

Key Takeaways
1Every page quoted on the fee question stated a specific percentage. None of the pages quoted on other questions stated a rate of its own.
2Two of the quoted pages put a figure in the first screen. No page in the other five had one there at all.
3Two pages sat at a dedicated fee web address, contained no percentage, and were not quoted on fees.
4The gap almost nobody fills: whether case expenses are deducted before or after the percentage.

Personal Injury AEO Across Answer Engines Starts With the Number on Your Fee Page

Open your firm’s fee page and read only the first screen. If there is no percentage in it, that is the cheapest fix on the site, and it is the one that separated the quoted pages from the rest more cleanly than anything else did.

Choosing well is harder than choosing to publish at all. Some figures belong on the page, some need a jurisdiction attached, and some cannot responsibly be stated. Lexicon writes to a rule that a number on a page has to be one a reader can verify, and the ones that cannot be verified get cut rather than hedged.

Call 877-486-8123 or contact us online and read us the first screen of your fee page, or look at Lexicon’s personal injury content work.

David Arato, JDs headshot

David Arato, JD is the founder of Lexicon Legal Content, an attorney-owned legal content marketing agency serving law firms since 2012. Lexicon writes attorney-reviewed content for law firms across criminal defense, personal injury, immigration, employment and family law. David is a frequent contributor to Attorney at Law Magazine and Attorney at Work and a recurring guest on legal marketing podcasts.

Criminal Defense FAQ Content Answer Engines Actually Cite

Posted on by David B. Arato

Google stopped showing FAQ rich results on 7 May 2026. Firms that read that as a reason to stop writing FAQ content read it wrong, and so did the ones who responded by rewriting their schema.

Answer engines kept quoting criminal defense FAQ pages straight through it. At Lexicon Legal Content we read every source link behind 131 cited URLs, and the pages that got quoted share one trait that has nothing to do with markup or with the word FAQ. Any firm tracking criminal defense FAQ answer engines by watching rich results has been watching a feature that no longer renders.

Q Short Answer

Answer engines quote criminal defense FAQ pages that put one question on its own URL, and they quoted none of the stacked FAQ pages tested.

None of the cited FAQ pages carried FAQPage markup, and FAQ rich results stopped appearing in Google Search on 7 May 2026.

The Rich Result Died and the Citations Did Not

Rich results and AI citations were never the same mechanism. Google’s structured data documentation confirms the FAQ search appearance is gone, with the rich result report and Rich Results Test support following it, after a 2023 change had already narrowed eligibility to government and health sites.

Citation runs on separate rails and it did not slow down. Law firm websites held 91 of the 131 URLs cited on criminal defense client questions, roughly 69 percent, with government and legal aid taking the rest.

Criminal defense firms are not trying to break into these answers. They are already inside them, so the useful question is what separates the pages that get quoted from the ones that do not.

What Gets Quoted Is a URL That Answers One Question

Seven cited pages sat in an FAQ directory, and every one of them answered a single question:

  • westcoastdefense.com/faqs/vehicle-search-with-a-warrant-in-california/
  • combswaterkotte.com/faqs/whats-the-difference-between-a-felony-and-a-misdemeanor-in-illinois/
  • davidmckenzielawfirm.com/faqs/what-happens-at-a-preliminary-hearing
  • brill-legal.com/our-services/criminal-defense/criminal-defense-faq/what-are-my-rights-if-i-am-questioned-by-police/

Across 131 cited URLs, not one stacked FAQ page, accordion or twenty-question list was quoted.

The reason is mechanical. An accordion holding twenty questions is one URL, so an engine answering one of them has no way to point at the part that matters. Twenty pages give it twenty places to point.

Two firms demonstrated what that buys. Combs Waterkotte was cited on the felony question and again on how long a case takes. David McKenzie Law Firm was cited on DUI reduction and again on preliminary hearings. Both from the same FAQ directory, one question per page, earning a second slot on a question unrelated to the first.

It Is Not the Markup and It Is Not the Label

None of the cited pages carries FAQPage structured data. Their markup is ordinary business schema: LegalService, Organization, LocalBusiness, WebPage, a BreadcrumbList. Whatever earned them a citation, FAQ markup was not it.

The word FAQ is not doing the work either. Pages filed under blogs were built identically and quoted just as often, including mccartylarson.com/what-happens-if-you-miss-a-court-date-in-texas/ and lafatalaw.com/blog/can-you-refuse-a-field-sobriety-test-in-illinois-what-to-know/.

What the cited pages share is narrower than a content type: one question, one URL, the question restated in the slug and again as the H1 in the words a client would use, usually with the state alongside it. Eighteen state and city names appeared inside cited slugs.

That puts the answer-first FAQ structure in its place. Format decides what an engine does after it reaches a page. Architecture decides whether it reaches it.

Expungement Is the Question That Is Not Open

Some criminal defense questions have no slot to win. Expungement returned zero law firm URLs, going to Illinois Legal Aid, Justia, Cook County courts, Michigan Legal Help, a federal court FAQ and Wikipedia.

Legal aid organizations own record-clearing content because record-clearing help is the service they exist to provide. Budget spent there competes with an incumbent that has owned the topic for years. The same budget spent on arrest, search or court-date questions lands where firms are already being quoted.

What to Change on a Criminal Defense FAQ Page

Three changes follow, and none of them is a new content program:

  • Split the accordion. Every question a client actually asks gets a URL, with the hub page kept as a directory linking out to them.
  • Say the question three times. In the slug, in the H1, and in the first sentence, phrased the way a client would type it, with the state named.
  • Stop spending on markup as the lever. No cited page relied on FAQ schema, and site structure for AI search is where the pattern actually appeared.

The last two cost an afternoon. Splitting costs real work, and it is the only one with a clear pattern behind it.

Frequently Asked Questions

Should a criminal defense firm split its FAQ page into separate URLs?

All seven FAQ pages cited in this testing answered one question on their own URL, and two firms were cited twice from the same FAQ directory. No stacked FAQ page was quoted at all.

Why do AI answers name different criminal defense firms each time?

Engines personalize to account activity and location. The same question asked signed in and signed out returned different lead firms in the same minute, so a signed-in visibility check reflects the searcher’s own history.

Key Takeaways
1Google stopped showing FAQ rich results on 7 May 2026, and answer engines kept quoting FAQ content anyway.
2Law firm websites held 91 of the 131 URLs cited across twelve criminal defense questions.
3All seven cited FAQ pages answered one question on their own URL, and none carried FAQPage markup.
4Split the accordion, give each question its own URL, and put the question and the state in the slug.

Criminal Defense FAQ Answer Engines Reward Structure, Not Volume

Splitting an accordion into separate pages is a week of work, and most firms already own the answers. What changes is where they live, which means the expensive part is architecture rather than writing.

Firms that would rather see their current citation list before rebuilding anything can start there. Call 877-486-8123 or contact us online and name the three questions your clients ask most, or read the law firm content writing we have produced for more than 300 firms across North America.

Criminal Defense GEO: Getting Recommended by ChatGPT, Claude and Gemini

Posted on by David B. Arato

There’s no such thing as getting recommended by AI.

There are four systems that answer the question differently, and when we asked them all for the best criminal defense lawyer in Phoenix on the same afternoon, ChatGPT and Google’s AI Mode came back with two lists that had nobody on them in common. Not one overlapping name.

So we stopped reading the answers and started reading the firms’ websites instead. That’s where criminal defense GEO across ChatGPT, Claude and Gemini turns into something you can act on.

Q Short Answer

No criminal defense firm was named by all four AI engines in either city we tested, but every firm that three of them agreed on states a state bar board certification on its homepage.

Only one of those four firms displays an aggregate review count, so review volume is not what these engines are reading.

AI Citations vs AI Recommendations for Law Firms

Citation and recommendation aren’t the same problem, and criminal defense firms tend to solve only one.

An engine cites you when it needs a source for a legal question. It recommends you when someone asks for a lawyer and it has to produce names. What these engines cite on criminal defense questions follows rules that don’t carry over to who they’ll name.

InterCore Research found a directory cited first in 77.8% of answers across 540 legal queries in July, on one engine, using phrasing that asks for a list. It held up on ChatGPT when we checked. It fell apart on Google.

Why No Criminal Defense Firm Made All Four Lists

Across Phoenix and Houston, on four engines each, the number of criminal defense firms named by all four was zero.

Four firms turned up on three of the four Phoenix lists. Two did in Houston. Nothing cleared all four, which is the first thing an AI search audit tends to surface.

That’s the finding a criminal defense firm should sit with before buying anything. Being the top result in one engine tells you almost nothing about the other three.

What the Four Most-Cited Criminal Defense Firms Share

Four criminal defense firms were named by three engines out of four: DM Cantor and Gurion Legal in Phoenix, DeGuerin & Dickson and Neal Davis Law Firm in Houston. Two states, one pattern.

We opened all four and checked the same seven things. Five showed up every time: a state bar board certification on the homepage with a badge, named attorneys with profile pages, practice areas split by charge type across dedicated pages, an FAQ, and the county named rather than just the city.

The sixth is interesting for being absent. Only DM Cantor shows an aggregate review count. The other three run individual testimonials and no total, which is awkward for the large part of this category selling review volume as an AI visibility lever.

Two of the four don’t just claim the certification, they price its scarcity. DM Cantor says no firm in Arizona has more board certified attorneys. Gurion Legal says fewer than a hundred attorneys in the state hold it.

Why Board Certification Beats Review Count

Here’s the mechanism, because a correlation without one isn’t worth acting on.

Board certification is the only claim on a criminal defense homepage that somebody else publishes. The State Bar of Arizona runs a public search of board certified attorneys with criminal law among the ten areas listed. The Texas Board of Legal Specialization does the same. Anyone can check in about forty seconds, and no amount of money moves you onto either list.

Everything else on that page is the firm’s word for itself. Years of experience, trial record, results, awards from properties that sell awards. An engine repeating any of it is repeating marketing copy and carries the risk of being wrong, which is the same reason outside verification does more for a law firm page than adjectives do.

Repeat a board certification and the worst case is that you’re accurate. That asymmetry is doing the work, and it sits comfortably inside what attorney advertising rules already allow you to claim.

The Criminal Defense Firms No Engine Named

A correlation among winners proves very little on its own, so we pulled four criminal defense firms that rank on page one for the same terms and were named by no engine at all.

Named attorneys, charge-type breakdowns, FAQ sections: nearly all of them had all of it. Those aren’t differentiators, they’re table stakes, and any advice built on them is advice to match the field.

Board certification was the row that split the two groups. Four of four among the firms that got named. One of four among the firms that didn’t.

That one exception matters and we won’t bury it. Orent Law Offices in Phoenix carries the certification prominently, plus named attorneys, an FAQ, charge segmentation, the county twice and a 4.9 rating from 54 reviews. No engine named them. The credential looks necessary and it clearly isn’t sufficient, so treat it as the floor rather than the plan.

Why We Run Every Engine Twice

A single clean Denver session in August showed Perplexity matching cited firms to the searcher’s city. It looked solid enough to write up.

A second session from Chicago returned Ohio firms, California firms, and on one query a full relocation to Memphis. One run is a hypothesis. Two runs are a finding, and the pages we’d already published were updated the same week.

That’s why every sentence below names the engine and the conditions rather than describing how AI works in general. Engine behaviour moves, and a single session will tell you almost anything you want to hear.

How ChatGPT Picks Criminal Defense Firms

ChatGPT cited Super Lawyers in Phoenix, then Expertise.com, Best Law Firms and Super Lawyers in Houston. That’s the complete list.

It explains why its shortlists looked nothing like anyone else’s. If those properties don’t have you, ChatGPT effectively doesn’t either, which is the practical case for treating directories and content as one system rather than a choice.

One correction worth passing to whoever runs your site. OpenAI documents four separate crawlers. OAI-SearchBot surfaces sites in ChatGPT search; GPTBot handles training. Blocking GPTBot does not take you out of ChatGPT results, and a lot of firms have blocked it thinking it does.

The rest of getting a law firm into ChatGPT is less technical than that.

Claude Argues With Its Own Sources

“Most of the pages that answer this question are pay-to-play directories or law firms ranking themselves.”

Claude wrote that, unprompted, before naming a single Phoenix firm. Then it pointed out that two of the top results were firms’ own best-attorneys listicles with themselves on the list, named State Bar of Arizona certification as the filter that actually matters, and sent us to azbar.org to verify it.

In Houston it did the same to Expertise.com and OnToplist.

No other engine behaved this way. On Claude, a credential somebody else publishes beats directory presence, and leaning on directory rankings gets a firm marked down by the engine reading them.

Gemini, Google AI Mode and the Three-Slot Problem

Gemini agreed with Google AI Mode more than it agreed with anyone else. In Phoenix the two shared four names; Gemini and ChatGPT shared one.

Gemini also has the least room. Testing across 1,200 buyer-intent prompts by Attrifast, which sells AI citation tracking, put its median at 2.4 citations per answer against 3.1 for ChatGPT and 6.4 for Perplexity. Three slots. Fourth place is nowhere.

AI Mode was the only engine leading with firm websites in both cities. It’s also running ads above the organic names now, four sponsored firms on the Phoenix query alone. Factor that into how local AI search strategy gets costed.

Frequently Asked Questions

Why Do Criminal Defense Firms Appear in Some AI Engines and Not Others?

The engines read different sources. In our August 2026 audit ChatGPT cited only directories, Google AI Mode cited firm websites first, and Claude cited directories while telling readers to distrust them.

Does Claude Recommend Law Firms Differently From ChatGPT?

Yes. Claude named firms and then argued against its own sources in both cities tested, directing readers to state bar certification records instead of directory rankings. No other engine did this.

Is Board Certification Necessary for Criminal Defense GEO?

Four of four firms that three engines agreed on state it. One of four firms nobody named does. It looks necessary and clearly is not sufficient, on a sample of eight firms in two cities.

How Many Sources Does Gemini Cite Compared With Other Engines?

Independent testing across 1,200 prompts put Gemini’s median at 2.4 citations per answer, the lowest of four engines measured, against 3.1 for ChatGPT and 6.4 for Perplexity. Three slots leaves no room for fourth place.

Key Takeaways
1All four firms that three engines agreed on state a state bar board certification on the homepage with a badge.
2Among four firms no engine named, only one states that certification, though nearly all had FAQs and named attorneys.
3ChatGPT cited only directories in both cities, while Google AI Mode cited law firm websites first in both.
4Put a credential a stranger can verify on your homepage, then check all four engines separately.

Where Criminal Defense GEO Across ChatGPT, Claude and Gemini Starts

Search your own firm name, then your city plus criminal defense lawyer, in all four engines. Takes twenty minutes. Write down which ones name you.

Then open your homepage and answer one question. Is there a credential on it a stranger could verify from a public record in under a minute? If not, fix that first, ahead of reviews and well ahead of schema.

The certification pages and charge-type coverage that change the answer are slower work, and that’s what our criminal defense content programs do for the 300-plus law firms across North America we’ve worked with. Book a strategy call and ask for the four-engine name check. We’re on 877-486-8123.

David Arato, JDs headshot

David Arato, JD is the founder of Lexicon Legal Content, an attorney-owned legal content marketing agency serving law firms since 2012. He built Lexicon a decade before any of these engines existed, and its research now tracks what they changed. David is a frequent contributor to Attorney at Law Magazine and Attorney at Work and a recurring guest on legal marketing podcasts.

What Immigration Firms Dominating Answer Engine Results Actually Do

Posted on by David B. Arato

A visitor to one of the most-cited immigration firm websites in Google’s AI answers is asked, in Spanish, whether they need the site in Spanish. As of 17 August 2026 the sentence was not a link, and the address it implies returned an error.

That firm holds the lead citation on whether an out-of-state immigration lawyer can represent a client. For the language a large share of its market speaks, it has a door painted on the wall.

Immigration firms competing for AEO and GEO visibility are being sorted by whether their Spanish content is a website or a widget. Across thirteen firms AI engines have already cited, five of the six appearing in Spanish answers run a genuinely separate Spanish site. Two of the seven appearing only in English do.

All thirteen were recorded as cited sources on Google AI Mode and Perplexity on August 2026, and their public pages were read and grouped by which language cited them. A translation tool leaving no visible trace would not show up in that reading.

Q Short Answer

The immigration firms appearing in Spanish AI answers run Spanish websites with their own addresses, while most of the firms winning the English answer have no Spanish pages at all.

Five of six firms cited in Spanish had a real Spanish section. Two of seven cited only in English did.

The Firms in the Spanish Answer Built Spanish Websites

Immigration firms appearing in Spanish AI answers run Spanish sites with their own addresses rather than translated versions of English ones. The two languages share no source domains at all, which is a fact about where the engines look rather than about which firms deserve finding.

Alonso and Alonso Law runs it backwards from everyone else, with Spanish at the root of the domain and English parked in a subfolder. It was the only firm here quoted in both languages on what representation costs.

Brown Immigration Law has the deepest Spanish section here, with more than ten Spanish blog posts carrying Spanish titles rather than translated English ones. Andrei Romanenko runs three languages, each at its own address, and Novo Legal and Vasquez Law both publish Spanish at addresses of their own. Five of the six firms cited in Spanish fit that description.

Most Firms Winning the English Answer Have No Spanish at All

The firms cited only in English mostly have not published Spanish, and several deliberately published something else. One lists multilingual staff in Vietnamese, Mandarin and French. Another runs a full second-language section that is Chinese. Those are decisions about who a firm wants to reach, made by practices that clearly know how to build one.

Two firms here did build Spanish properly, one across six languages and one as a complete parallel tree. Then there is the firm from the opening, whose Spanish offer goes nowhere. It is not absent from AI answers. It leads one of them.

A Spanish URL Is Not a Spanish Page

The most expensive mistake in this set is a Spanish address wrapped around English writing, because it looks solved from the dashboard. One firm runs ten languages through a translation layer, and on August 2026 its Spanish section put Spanish menus above articles still written in English. Every blog title listed on that Spanish page was in English. A reader arriving in Spanish gets a Spanish menu and then hits a wall.

An engine assembling a Spanish answer wants Spanish sentences it can quote, and menu labels are not sentences. The furniture got translated and the content did not, which is what happens when language is treated as a plugin setting rather than as writing someone has to commission. The firm pays for the layer and still has nothing quotable.

The Pattern Breaks in Both Directions

Two firms contradict the pattern, and both matter before anyone treats Spanish pages as a switch. One appeared in a Spanish answer with no Spanish content anywhere on its site, in a placement marked Sponsored. Running the other way, the two English-cited firms with genuine Spanish sections turned up in no Spanish answer here.

So a Spanish site is neither necessary nor sufficient. The evidence supports something narrower: firms cited in Spanish overwhelmingly have real Spanish pages, and thirteen sites is a sample rather than a study.

Nobody Has Measured This Market in Its Own Language

The gap between what immigration clients speak and what anyone measures is documented at the source. Census Bureau figures for 2020 to 2024 put 22.3% of US residents aged five and older speaking a language other than English at home, and on the 2018 to 2022 figures 61.1% of that group spoke Spanish. Published research has not gone near it: InterCore Research ran 540 legal queries through Perplexity in July 2026 on three fixed phrasings, all English.

Frequently Asked Questions

Does a translation plugin count as Spanish content for AI search?

Only when it produces Spanish sentences. A plugin translating menus while leaving article bodies in English gives an engine nothing quotable, which was the case on one immigration firm site read in August 2026.

Do a law firm’s Spanish pages need their own web addresses?

Every firm cited in Spanish here had Spanish content at a distinct address, whether a subfolder or the root domain. Content appearing only after a visitor clicks a switcher has no address for an engine to cite.

Is a bilingual tagline enough to appear in Spanish answers?

No. Phrases like “Se Habla Español” describe the practice rather than answer a question, and the firms cited in Spanish paired that messaging with actual Spanish pages. One firm read in August 2026 offered Spanish in its header with nothing behind it.

Key Takeaways
1One firm cited in Spanish runs Spanish as its main site, with English as the translated version.
2A Spanish menu wrapped around English articles is a common and expensive halfway house.
3The pattern has exceptions in both directions, so Spanish pages are not a guarantee of anything.
4Open your own Spanish page, check it has its own address, and read the article underneath it.

How to Check Your Immigration Firm’s AEO and GEO Visibility

Checking immigration firms for AEO and GEO visibility across two languages takes about ten minutes. Open your own Spanish page and confirm it has its own web address rather than appearing in place when a switcher is clicked. Then open a Spanish article and read the body, not the menu. If the words underneath are English, the engines are reading English too. Do the same on the three firms you lose work to.

Knowing which of your pages actually exist in each language is the step most firms skip, and it decides what to build before anyone writes a word. Call 877-486-8123 and tell us which questions your clients ask, or see the content built to be quoted first. You can also contact us and we will read your site in both languages and tell you what is really there.

David Arato, JDs headshot

David Arato, JD is the founder of Lexicon Legal Content, an attorney-owned legal content marketing agency serving law firms since 2012. Lexicon has published competitive analyses of AI search visibility across immigration, employment, family law and criminal defense. David is a frequent contributor to Attorney at Law Magazine and Attorney at Work and a recurring guest on legal marketing podcasts.

What AI Actually Cites on Criminal Defense Questions

Posted on by David B. Arato

Somebody’s brother got arrested tonight. It’s eleven o’clock, they’re sitting in a car outside the county jail, and they’re typing “how much does a criminal defense lawyer cost” into a phone.

Google’s AI answered that question for us last week using two law firms. Not Avvo, not Justia, not Super Lawyers. A four-attorney shop in Maryland and a firm in Dallas, each one cited off a single page.

We read both pages. What they have in common is short enough to fix this quarter, and it isn’t what anyone selling criminal defense AI citations is telling you to do.

Q Short Answer

Two small law firms hold Google’s answer on what a criminal defense costs, and the pages that won it name their state in the H1, price by charge class, and commit to real numbers.

One of those pages runs 5,000 words and the other runs 650, so length is not the variable.

Which Criminal Defense Pages Google’s AI Cited

Scheuerman Law is in Maryland. Their cost page runs about 5,000 words, opens with a quick-answer block that puts a dollar figure in the first sentence, and carries a fee calculator, a ten-question FAQ, a named author and a sources section at the bottom. It’s the page every content marketer would tell you to build.

The Gallian Firm is in Dallas. Their version runs 650 words, has no FAQ, no named author, no answer block, and opens by saying the answer depends on several factors.

Google cited both. Same answer, same day.

So length didn’t do it. Neither did the calculator, the byline, the FAQ or the answer-first structure, because one page has all of that and the other has none of it and they both got quoted.

5 Things Both Cited Pages Share

Both pages do five things, and that’s the whole list.

Both put the state in the H1. Both give specific dollar figures rather than hedged ranges, and both attach those figures to charge classes, so a Class B misdemeanor has its own number and a first-degree felony has a different one.

Both explain flat fee versus hourly in a section of its own. And both pull DUI out and price it separately.

Here’s the mechanism, and it’s the part the tips never explain. The engine was asked for a number, and almost every page on that SERP refuses to give one, because giving one is uncomfortable and every lawyer’s instinct is to say it depends.

When a question wants a figure and only two pages will commit to a figure, those two get quoted. Reluctance is what’s clearing the field, and it’s the same opening that shows up across blog content built for AI citation.

We wouldn’t spend another dollar on schema markup until your cost page names your state and prices a Class B misdemeanor.

How Question Type Changes Which Sites Get Cited

Criminal defense questions sort into three groups, and each group hands its citations to a different kind of site. Nine questions, four engines, one afternoon in August.

Ask with a city and an intention to hire, something like “best criminal defense lawyer in Phoenix,” and you get named firms with their own websites cited underneath.

Ask what a defense costs and you get no recommendation at all, but firm content pages carry the citations. That’s the group Scheuerman and Gallian won, and it lines up with how client search journeys split before anyone types a firm name.

Ask how to choose a lawyer in the first place and law firms vanish.

The One Criminal Defense Question No Firm Wins

“How do I find a good criminal defense attorney” returned several hundred words from Google’s AI and not one law firm. The sources were Super Lawyers, a Reddit thread from 2021, two YouTube channels and an awards directory.

Nothing on your website is going to take that slot this year. The engine has decided the honest answer to how-do-I-choose comes from somewhere other than the people being chosen, and it’s hard to argue the logic. Directories hold that ground, which is its own argument for how the two work together.

Knowing that is worth real money, because it’s the query most firms would instinctively target first. Skip it. Spend the budget on the cost question, where two small firms just proved the door is open. That’s the difference between generative engine optimization and pattern-matching a keyword list.

What AI Visibility Advice Leaves Out

We checked fourteen pages ranking for criminal defense AI visibility terms this month, and thirteen were agency service pages written to look like analysis.

None ran a query. None named a domain that got cited. None linked a primary source for a single number.

Most of them sell schema markup as the lever. Google’s own guidance on generative AI features says structured data isn’t required for these features and there’s no special markup to add. Schema still earns rich results in classic Search. Different argument, different budget line.

We run the same check on our own pages, and this piece turned one up. A page on this site credits Bain with a figure about how many people research legal questions through AI tools, and that number isn’t in the Bain release. It’s being updated.

We mention it because the test we’re describing only means something if it applies in both directions. Hold whoever writes your content to it.

The scale is real even when the stats around it aren’t. Ahrefs put AI Overviews on 23.6% of legal SERPs across 146 million results, and Pew found Google users click a link far less often when an AI summary sits above them. Being the source inside the summary is the traffic now.

Frequently Asked Questions

Do AI Engines Cite Criminal Defense Law Firm Websites at All?

Yes, on some question types. In our August 2026 audit, firm websites carried the citations on city-specific hiring questions and on the cost question, and were absent entirely from general how-to-choose questions.

Which Criminal Defense Questions Are Hardest to Win?

General questions with no city and no charge type. “How do I find a good criminal defense attorney” returned directories, a Reddit thread and YouTube videos, with no law firm cited anywhere in the response.

Does Schema Markup Drive Criminal Defense AI Citations?

Google’s AI optimization guidance says plainly that structured data is not required for generative AI features and that no special markup exists for them. Schema is still worth implementing for rich results in classic Search.

How Often Should a Criminal Defense Firm Re-Check Its AI Visibility?

Quarterly. Engine behaviour moves faster than that, but a quarterly check catches drift without turning into a research project, and anything you find in one run should be re-run before you act on it.

Where Criminal Defense AI Citations Start

Open the page on your site that answers what a criminal defense costs. If you don’t have one, that’s the finding.

If you do, check three things. Is your state in the H1. Is there a dollar figure a person could repeat out loud. Does a Class B misdemeanor have a different number from a second-degree felony.

Fix those three and you’ve done more for your criminal defense AI citations than a year of schema work. If you’d rather see the whole picture first, that’s what an AI search audit is for. When you want the pages built properly, every piece of criminal defense content we deliver is screened against ABA Model Rule 7.1 before it leaves the building. Book a strategy call or reach us on 877-486-8123, and ask for the cost-page check.

David Arato, JDs headshot

David Arato, JD is the founder of Lexicon Legal Content, an attorney-owned legal content marketing agency serving law firms since 2012. Lexicon was writing legal content for a decade before answer engines started deciding which pages get quoted. David is a frequent contributor to Attorney at Law Magazine and Attorney at Work and a recurring guest on legal marketing podcasts.

Family Law GEO: Getting Recommended by ChatGPT, Claude and Gemini

Posted on by David B. Arato

Two family law firms run the same playbook. Both publish jurisdiction-specific pages, both structure answers cleanly, both get their statutes right. One gets named when a searcher asks who to hire, and the other does not.

Lexicon Legal Content asked ChatGPT, Claude and Gemini the same six California divorce questions on the same day in August 2026. The three engines did not disagree at the margins. They behaved like three different products.

Q Short Answer

In the Lexicon Three-Engine Family Law Test (August 2026), ChatGPT cited law firm websites on 0 of 6 California divorce questions, Claude on 4 of 6, and Gemini on 6 of 6. Only Gemini named a law firm inside the answer text, on 6 of 6 questions.

Being cited and being named are different results. In this test, one engine out of three produced the second one.

How this was tested

The Lexicon Three-Engine Family Law Test

EnginesChatGPT, Claude, Gemini
Questions6, covering property division, custody, self-representation, uncontested divorce, cost, and how to hire
JurisdictionCalifornia, named in every question
RecordedEvery cited domain, and whether any law firm was named in the answer text
DatesAugust 2026
LimitsOne state, one day. California community property rules may not generalize. The Gemini session used a business Google account.

What did each AI engine cite for California divorce questions?

Each engine produced a different result on identical questions. The Lexicon Three-Engine Family Law Test recorded the following across six California divorce questions in August 2026.

EngineCited law firm sitesNamed a firm in the answer
ChatGPT0 of 60 of 6
Claude4 of 60 of 6
Gemini6 of 66 of 6

ChatGPT produced 29 citations across the six questions and sent every one to the California courts self-help site or the State Bar. Claude searched on four questions and law firm content dominated what it found, reaching 6 of 8 sources on the divorce cost question. Gemini grounded all six answers and attributed claims to firms by name in the visible text.

What is the difference between being cited and being recommended by AI?

A citation is a link to a page in the sources panel. A recommendation is a firm named inside the answer as somewhere to look, which can happen with or without a link.

The Lexicon Three-Engine Family Law Test found these are separate outcomes rather than degrees of the same one. Across 12 ChatGPT and Claude answers, zero law firms were named in the answer text. Across 6 Gemini answers, a firm was named every time.

Firms are being sold answer engine optimization packages that promise the second outcome and perform work that produces the first. The work is real. The promise runs ahead of it.

Does Claude search the web for family law questions?

Not always. Claude answered 2 of 6 questions in the Lexicon Three-Engine Family Law Test without searching at all: property division and child custody.

Both answers were accurate, citing California Family Code section 3011 from memory, with no sources panel and nothing linked. No page could have been cited on those questions regardless of how well it was built, because the engine never went looking.

Gemini grounded on those same two questions, so this reflects how one engine handles settled state law rather than a general property of AI search. The practical consequence stands either way: part of what an engine says about a practice area is fixed before a firm’s content enters the picture.

Which law firms did AI engines cite most often?

Four firms appeared on two engines independently in the Lexicon Three-Engine Family Law Test: Griffith Young, Moshtael Family Law, Provinziano & Associates, and Erica Bloom Law.

Two engines with separate retrieval systems, run separately, surfaced the same firms. That is not one product’s ranking quirk. Those firms have published content on the specific questions people ask, at a level of detail that survives being summarized.

Cross-engine repetition is the most useful pattern in the dataset, because it is the part that does not depend on any single engine’s behavior. It is what the signals engines use look like in practice rather than in theory.

Do AI engines recommend law firms to hire?

Rarely, and never straightforwardly. On the hiring question, all three engines in the Lexicon Three-Engine Family Law Test routed to institutions rather than to law firm marketing.

ChatGPT cited the California courts and the State Bar. Claude cited seven sources and all seven were bar referral services. Gemini named a firm, then wrapped it in the State Bar attorney search, county bar referral panels and third-party directories, and told the reader not to choose off sponsored ads.

The most commercially valuable query produced the most consistent refusal. A firm competing for that answer is not competing with other firms, it is competing with the State Bar, and the route in is being the content the institutional answer draws on.

How should a family law firm measure AI visibility?

Run the same questions on more than one engine and record two things per answer instead of one. Which domains were cited, and whether any firm was named in the answer text.

The second column is the one almost nobody records, and it is the only way to tell a citation problem from a recommendation problem. Our own California audit took under two hours.

Three outcomes, three different responses. If firms are cited but yours is not, that is a content problem and content work moves it. If the engine answers without searching, no content changes that query. If firms appear on one engine and nowhere on another, that is the engines being different products rather than a failure, and the fix is measuring more than one.

Most family law firms have never checked which of the three they face. The urgent, high-stakes questions are worth testing first, because that is where a searcher decides whether to call anyone at all.

Frequently Asked Questions

Does ChatGPT recommend law firms?

No, not in the Lexicon Three-Engine Family Law Test (August 2026). All 29 ChatGPT citations across six California divorce questions went to the California courts self-help site or the State Bar, and no law firm was named in any answer. That included the question asking directly how to find a divorce lawyer.

Does Claude cite law firm websites?

Yes, on informational questions. In the Lexicon Three-Engine Family Law Test, Claude searched on 4 of 6 California divorce questions and law firm content made up most of what it found, including 6 of 8 sources on the divorce cost question. On the hiring question, Claude cited seven sources and all seven were bar referral services.

Does Gemini name law firms in its answers?

Yes. In the Lexicon Three-Engine Family Law Test, Gemini named a law firm inside the answer text on 6 of 6 California divorce questions, including Szeto-Wong Law, Provinziano & Associates, Moshtael Family Law and Candelaria LLP. ChatGPT and Claude named zero firms across their twelve combined answers.

What is the difference between being cited and being recommended by AI?

A citation is a link to a page in the sources panel, produced when an engine searches and finds it. A recommendation is a firm named inside the answer as somewhere to look. The Lexicon Three-Engine Family Law Test found 12 ChatGPT and Claude answers naming zero firms while 6 Gemini answers named one every time, which makes these separate outcomes that need separate measurement.

How should a family law firm measure AI visibility?

Pick six questions clients actually ask, name the state in each, and run them on at least two engines. The Lexicon Three-Engine Family Law Test recorded two columns per answer: which domains were cited, and whether any firm was named in the answer text. There is no analytics report for the second one, because a firm named without a link produces no referral traffic, so it has to be observed directly rather than estimated.

Key Takeaways
1All 3 engines declined to send searchers to a law firm’s own site on the question about how to hire a lawyer.
2Claude answered 2 of 6 questions without searching, so no page could have been cited on either one.
34 law firms appeared on 2 engines independently: Griffith Young, Moshtael Family Law, Provinziano & Associates, and Erica Bloom Law.
4Testing one engine tells a firm nothing about the other two.

Before You Buy an AI Visibility Package

Any vendor quoting a single AI visibility score is averaging across engines that behaved nothing alike on identical questions. That number cannot identify which of the three problems a firm has, and the three need different work.

Every piece Lexicon delivers is reviewed by a licensed attorney before it reaches the client. Call 877-486-8123 with the six questions clients ask most, or start with the legal content writing. You can also contact us with what you are seeing.

David Arato, JDs headshot

David Arato, JD is the founder of Lexicon Legal Content, an attorney-owned legal content marketing agency serving law firms since 2012. Lexicon tests AI engines directly and records whether a firm is named in the answer, not just which pages get linked, because those turn out to be different results. David is a frequent contributor to Attorney at Law Magazine and Attorney at Work and a recurring guest on legal marketing podcasts.

What Employment Law Firms Dominating Answer Engine Results Actually Do

Posted on by David B. Arato

Five sessions across two engines and two states have now returned the same Dayton law firm’s blog post when we asked how a person can tell whether a firing was illegal. By most published checklists that page should not be winning anything.

At Lexicon Legal Content we spent August testing which employment law firms answer engines actually cite, and the pattern in the pages that keep surfacing is not the one the industry keeps recommending.

Q Short Answer

The employment law firms getting cited by answer engines are small practices publishing plain explanations of client questions, not national brands and not firms with the best-optimized service pages.

Citations split across 254 firm websites in one study, so no incumbent has to be displaced for a firm to enter the set.

The Firms Winning Employment Answers Are Ones You Have Not Heard Of

Duwel Law, a Dayton practice, has been cited five times across Google AI Mode and Perplexity, from sessions run in Colorado and Illinois, on the question of how to recognize wrongful termination. Nisar Law in New York came back on hostile work environment questions in run after run, and again on damages. Bell Law Group and Amber Boyd Law both reappeared weeks apart on different engines.

Not one of them is a national brand. There is no Am Law firm in the set, and the only large firm we saw all month showed up on a directory-flavored question rather than a client one.

What the four have in common is narrow and repeatable. Each is a practice focused on employment work rather than a general firm with an employment page. Each publishes explanatory blog content on the questions clients ask, not service copy describing what the firm offers. And each states its geography plainly enough that an engine never has to guess where the advice applies.

That is not an accident of our sample. InterCore Research, which ran 540 high-intent legal queries across 30 US metros in July 2026, found firm citations split across 254 different websites, with the most-cited single firm appearing 36 times out of 18,900 citations. Nobody owns this category, which means nobody has to be displaced.

What a Cited Employment Page Actually Looks Like

We went back and read the post engines keep quoting. It is a blog post from March 2026, roughly 1,200 words, and it breaks most of the rules our industry publishes.

There is no named attorney byline, only the firm name. It carries no external citations, no FAQ section, and it opens on how losing a job feels rather than answering the question in its own title.

What it has is a page a machine can take apart. Short labelled sections, bolded key terms, three scannable lists, a five-step numbered sequence at the end, and its city and five surrounding counties stated without ambiguity.

Duwel Law hits one of the five signals on our own list of what engines check, and gets cited anyway. That is not the list failing. It is the list mattering less on a question where the reader wants an explanation than on one where they are choosing who to call.

Extractability wins the informational questions. Author and credibility signals carry the commercial ones, where a person is deciding who to trust with a claim.

They Win the Questions Directories Cannot Answer

InterCore found that 77.8% of answers cited a directory first, asking variations of “top employment law firms in Chicago.” Across the sixteen answers we collected asking client-phrased questions, a directory led exactly none of them.

Both findings are correct, and the gap between them is where these small firms live. Justia and Super Lawyers were built to answer “who should I hire in my city” and they win it decisively. They have almost nothing to say about whether your boss can fire you for no reason, what counts as a hostile work environment, or what your case might be worth.

Duwel Law is not out-competing Justia. It is answering a question Justia never tried to answer. That distinction is worth more than any single tactic in answer engine optimization.

The practical version for a firm owner: stop writing pages that describe your employment practice and start writing pages that answer the questions a person has before they know they need a lawyer. The firms getting cited did not out-spend anyone. They wrote the explanation first.

Being Named Is Not the Same as Being Cited

There is a version of this that looks like a win and is not. The firm appears in the answer, but the firm’s own website is nowhere in the sources.

InterCore found 34.2% of firms named in an answer had no page of their own among the cited sources. The engine knew those firms only from directory listings, which means the description a prospective client reads was written by Avvo, and there is no link back to anything the firm owns.

The firms in the first section are not in that position. They are cited because a page they wrote is doing the work.

Why You Probably Cannot See Any of This From Your Own Account

Here is the part that makes this hard to verify for yourself. We asked Perplexity six employment law questions from a Chicago connection, then asked again after turning off the account history that had been saving our searches.

On “do I need a lawyer for a wrongful termination case,” both runs returned eight law firms, and exactly one firm appeared in both answers. A third run, with conditions held steady, came back identical to the second on two of four questions. These answers are reproducible, which means the difference between the first two runs was us.

A firm checking its visibility while signed in to the account it uses for competitor research is reading its own history back to itself. Location matters less than most people assume. From that Chicago connection the same question returned Ohio firms once and California firms twice, and never an Illinois practice.

None of this is stable, and any firm treating a single good result as an achievement should know that. Duwel Law sat at position 2 in one of our runs and position 13 in another, on the same question, thirty minutes apart. Consistent citation across sessions is the thing worth chasing. A single appearance is closer to weather.

Where to Put the Next Page

In every run we did, usa.gov or the EEOC held the first citation slot on definitional and procedural questions. That ground is not available, and no amount of content investment will take it.

What sits unclaimed is everything the agency does not publish. What a charge actually costs, how long an EEOC process runs in practice, whether a particular set of facts is worth filing on, what happens to your job while a charge is pending, and what your state adds on top of federal law.

Build those as guides rather than short posts and connect them, which is what topic cluster strategy is for. Otterly.ai, analyzing more than a million cited URLs in 2026, found guide pages averaged 2.7 citations against 2.0 for blog posts.

What Not to Spend On

One in five employment law sites carries no structured data, according to a March 2026 audit of 299 employment law websites by AISO Studio, a company that sells AI visibility auditing. Duwel Law’s cited page has no FAQ schema either, and it is cited anyway.

Google’s own guidance states that structured data “isn’t required for generative AI search, and there’s no special schema.org markup you need to add.” Schema still earns rich results in classic search, which is a separate and legitimate argument.

Nor does a top ten ranking carry a citation the way it once did. Ahrefs put AI Overview citations coming from the top ten at 37.9% in March 2026, down from 76% the previous July, though the same authors note they improved their citation parsing in between, so part of that drop is method rather than behavior.

How to Check Whether You Are in the Set

Sign out first, or use a private window. Ask the six questions your clients actually ask, in their words rather than a directory’s, and read the source list under each answer.

Do it twice on different days and write down the date and whether you were signed in. Six questions takes twenty minutes, and one reading on its own tells you very little.

Frequently Asked Questions

Which employment law firms are getting cited by AI search?

Small and mid-size practices focused on employment work. In our testing Duwel Law, Nisar Law, Bell Law Group and Amber Boyd Law recurred across engines and sessions, while national brands appeared only on directory-style questions.

Do I need a big budget to get cited by answer engines?

No. Citations split across 254 firm websites in InterCore’s study, with the most-cited firm appearing 36 times out of 18,900. The category has no incumbent, so the barrier is content structure rather than spend.

Why does being signed in change which law firms an engine cites?

Engines personalize to account activity. A firm that has spent months researching competitors from one account will see that history reflected back, which makes its own visibility check the least reliable one available to it.

Do legal directories dominate AI answers or not?

It depends entirely on the question asked. Directories led 77.8% of answers to “top firms in city” phrasing in InterCore’s testing, and none of the sixteen client-phrased questions that we ran ourselves.

Key Takeaways
1One Dayton firm’s blog post was cited in five sessions across two engines and two states.
2That page has no attorney byline, no external citations and no FAQ section.
3Directories led 77.8% of answers to city-and-practice queries and none of the client-phrased ones we ran.
4Write the explanation a client needs before they know they need a lawyer, and structure it to be quoted.

The Harder Half

Twenty minutes gets you the reading. Knowing which gaps are worth a page and which belong to a federal agency you will not outrank is the harder half, and it is where we start. Every piece we deliver is reviewed against the statute it cites before it reaches you, which matters more in employment law than in most practice areas, because the wrong sentence about a filing deadline is not a marketing problem.

Call 877-486-8123 and tell us which questions your clients ask most often, or see the legal content writing itself first. You can also contact us with what you are up against.

David Arato, JDs headshot

David Arato, JD is the founder of Lexicon Legal Content, an attorney-owned legal content marketing agency serving law firms since 2012. He has spent the past year rebuilding how Lexicon measures AI visibility, after finding that most published methods produce numbers a firm cannot act on. David is a frequent contributor to Attorney at Law Magazine and Attorney at Work and a recurring guest on legal marketing podcasts.

How Long Until a Law Firm Appears in AI Search?

Posted on by David B. Arato

Twelve pages currently answer this question with a timeline. Ten of them sell the service whose duration they are estimating, and not one of those ten cites a source for the number it gives you.

That number, three to six months, has turned up in nearly every content audit that crosses our desk at Lexicon Legal Content, and we have never found the study behind it. There is no study. It is a retainer length, not a finding.

Q Short Answer

There is no waiting period: a law firm page becomes eligible for AI citation as soon as it is indexed, not after a fixed number of months.

The three to six month figure repeated across the pages ranking for this question is a retainer length, not a research finding.

Nobody Ranking for This Question Sources Their Answer

The three-to-six-month figure traces back to nothing. MarGen quotes three to six months, with compounding results from six to twelve, on a page selling packages that run £2,950 to £12,950 per month, and that page carries no citations at all. RankBrain puts first citation at 30 to 45 days while selling retainers of $2,500 to $10,000.

Site Your Case and Argota repeat the same three-to-six-then-six-to-twelve ladder, also unsourced, and both sell law firm SEO. Only Semrush and DerivateX cite anything, and DerivateX is citing Semrush.

The ladder is inherited from pre-AI SEO retainer copy, which is where those durations have always come from. If you are deciding where a small budget goes, the duration you are quoted is the product being sold, not evidence about the product.

There Is No Waiting Period, and Both Platforms Say So

Eligibility is immediate. OpenAI’s publisher documentation states that any public website can appear in ChatGPT search, and Google’s guidance on AI features says there are no additional technical requirements and no special schema markup needed to qualify.

Neither company names a minimum site age, an authority score, or a traffic threshold. Google’s actual condition is narrower and far more practical than tenure: the page must be indexed and eligible to be shown in Google Search with a snippet.

So the question to ask is not how long you wait. It is whether you are indexed, and after that, whether the page does what engines actually weigh when they choose a source to quote.

Crawled, Indexed, and Cited Are Three Different Events

Most timeline advice collapses three separate events into one number. A page has to be crawled, then indexed, then selected as a citation, and only the first of those three has any published guidance attached to it.

Crawling alone can take anywhere from a few days to a few weeks, a range Google publishes in its instructions for recrawling your URLs, which also warn that requesting a crawl does not guarantee that inclusion in search results will happen instantly or even at all.

Perplexity documents its crawlers but publishes no crawl frequency and no citation eligibility rules. OpenAI publishes no refresh frequency at all. Any single number covering all three events is filling two gaps the platforms themselves have left open.

The One Real Experiment Found Something Faster and Less Stable Than Anyone Is Selling

The only controlled test on record found first citation far faster than the sales copy claims, and retention far worse. Semrush published 81 FAQ-style pages and queried both engines daily for 30 days.

Google AI Mode cited 29 of the 81 pages, 36 percent, the day after publication, and 45 pages, 56 percent, by day seven. By day 30, only 21 pages, 26 percent, were still being cited. ChatGPT moved in the opposite direction, starting at 8 pages on day one and climbing to 34, or 42 percent, by day 30.

Semrush read its own Google numbers this way: the engine seems to be continuously re-evaluating which sources to cite, adding and removing pages from its pool of trusted sources on an almost daily basis.

The caveat matters as much as the finding. Semrush ran this on its own blog and wrote that the pages were given the best possible conditions to succeed, which makes it a measurement of new pages on an already-authoritative domain. It does not measure a new firm’s new site, and no published study does.

Appearing Once Is Not Arriving

Holding a citation is a different problem from earning one, and it is the deeper reason any single date is the wrong thing to ask for. Kevin Indig’s July 2026 analysis of 1,094 categories and more than 600,000 citations found the leading source switched in 1,950 of 5,470 month-over-month comparisons.

Where an entrenched owner exists, the picture reverses. A clear owner held first place in 90.4 percent of those comparisons, and the difference sits in the margin: categories that changed hands had a median lead of 1.3 points, against 2.9 points for the ones that held.

For a small firm, the opening is not displacing an owner. It is the narrow-margin categories, and in legal specifically Indig found no clear owners at all, which is why whether a solo can compete turns on turnover rather than firm size.

Frequently Asked Questions

How long does it take for a law firm to show up in ChatGPT?

No vendor timeline survives checking, so build your own baseline instead. Record which prompts name your firm today, then re-run those same prompts monthly. Movement on that list is the only date worth tracking.

Do I need to wait until my site has authority before AI will cite it?

No. Neither platform publishes an authority threshold or a minimum site age. What decides a citation is whether the page answers the question in a passage an engine can lift, and a new site can do that immediately.

Why did my firm appear in an AI answer once and then disappear?

One appearance is a sample of one query on one day. Re-run the same prompt several times, and across more than one engine, before drawing a conclusion, since results shift between sessions even when nothing on your site changed.

Is AI search optimization different from SEO?

Less than vendors suggest. Both depend on indexation and on answering clearly. The working difference is that AI answers quote passages rather than rank pages, so self-contained paragraphs matter more than keyword placement does.

Key Takeaways
1Ten of the twelve pages ranking for this question quote a timeline while selling the service, and cite no source for it.
2OpenAI and Google both state that eligibility depends on indexation, not on site age, authority score, or traffic.
3Semrush measured 36 percent of new pages cited by Google AI Mode within 24 hours, and only 26 percent still cited at day 30.
4Track which prompts name your firm and re-run them monthly, because holding a citation matters more than the date you first earned one.

Want the Version of This With Your Own Pages in It?

Ask us which of your pages an engine can currently quote, and we will tell you what is blocking the ones that cannot be. Every piece we deliver at Lexicon Legal Content is checked back to its primary source before it leaves us, then reviewed by a licensed attorney admitted in Missouri and Illinois.

Call 877-486-8123 or contact us online and ask for the citation check.

David Arato, JDs headshot

David Arato, JD is the founder of Lexicon Legal Content, an attorney-owned legal content marketing agency serving law firms since 2012. He has watched law firms buy timelines rather than outcomes for most of that period, and built Lexicon’s research process around tracing every number back to its primary source. David is a frequent contributor to Attorney at Law Magazine and Attorney at Work and a recurring guest on legal marketing podcasts.

Can a Solo Attorney Compete With Big Law in AI Search?

Posted on by David B. Arato

In a study of 18,900 citations drawn from 1,620 AI answers about lawyers, the most-cited single law firm website appeared 36 times. Big Law is not winning AI search either, which turns the question from whether a solo can catch up into what is actually being cited instead.

Those numbers come from InterCore Technologies, published 20 July 2026 across 540 queries run three times each. InterCore sells visibility services, and it publishes its full query panel and its raw per-prompt data, which is more than most research in this space offers.

Q Short Answer

Yes, because Big Law is not winning AI search either.

In a study of 18,900 citations about lawyers, the most-cited single law firm website appeared 36 times, and legal directories took more than half of everything else.

The Firm You Are Losing To Is Not a Firm

Legal directories took at least 51.8 percent of every citation in the State of AI Search Visibility dataset and were the first source cited in 77.8 percent of answers.

The eight most-cited domains were all directories before a single law firm appeared. Justia took 9.7 percent, Super Lawyers 8.07, Avvo 5.74, Lawyers.com 5.72. The 254 law firm websites cited at all divided what remained.

The finding that should reorganize your thinking is this one: 34.2 percent of the firms these systems named by name had no page of their own among the cited sources. The engine knew those firms entirely from someone else’s listing, which is the practical case for understanding how directories drive AI visibility before touching anything else.

Site Authority Barely Predicts Whether an AI Cites You

Across roughly five million citation source URLs, the correlation between domain authority and citation frequency was close enough to zero to be treated as noise.

That analysis ran 20,000 prompts across AI Mode, AI Overviews, ChatGPT and Perplexity over three months, and Surfer published what it found about domain authority and AI citations on 14 July 2026. PageRank correlated at negative 0.065, its Domain Score at negative 0.051, and Harmonic Centrality at positive 0.010. The study bucketed sources by citation frequency and re-ran with the strongest five percent of domains removed, and the coefficients stayed flat.

Surfer sells a tool that benefits from that conclusion, which is worth holding in mind while you read it. The sample and the controls are disclosed, which is what makes the number usable.

Ranking on Page One Is Not the Entry Ticket

Only 37.9 percent of AI Overview citations come from pages ranking in Google’s top ten.

That figure comes from 863,000 keyword result pages and four million Overview URLs, published on 2 March 2026 in Ahrefs’ count of AI Overview citation sources. It supersedes an earlier 76 percent that sampled only the three most visible citations per answer, and Ahrefs attributes part of the change to its own improved citation parsing rather than to Google alone. Both numbers are accurate about different things, and only one of them describes your odds at earning AI Overview citations in an ordinary slot.

The mechanism behind it is documented. AI Overviews and AI Mode “may use a ‘query fan-out’ technique, issuing multiple related searches across subtopics and data sources,” reads Google’s account of how AI features select sources, which adds that these features “offer unique opportunities for more types of sites to appear.” Assembling documents responsive to related queries rather than only to the one typed is also what the patent for generative summaries describes. That fan-out specifically advantages small sites is a reasonable inference from this, not a documented finding.

The only independent probability sample in this evidence comes from Pew Research Center, whose count of who AI summaries actually cite found Wikipedia, YouTube and Reddit together accounting for 15 percent of sources in AI summaries against a comparable 17 percent in ordinary results. AI answers are not meaningfully more concentrated than search already was.

What This Does Not Mean

None of this means a small site wins by default.

Only 36 brands held visibility across every platform in Semrush’s 2026 AI Visibility Index, published 26 June 2026 from 126 million prompts, which also found Gemini averaging three sources per response against ChatGPT’s fifteen. A three-source answer is a far tighter funnel than a fifteen-source one. Category owners held position in 90.4 percent of month-over-month comparisons once established, on Kevin Indig’s July 2026 read of whether topical authority matters across more than 600,000 citations, against 89.3 percent of AI search demand still sitting in categories with no clear owner at all.

Both readings hold. Incumbency is sticky where it exists, and in most legal categories it does not exist yet.

What a Solo Should Actually Do About It

Control the directory profiles already speaking for you, then write for the specific questions rather than the head term.

“Any public website can appear in ChatGPT search,” reads OpenAI’s publisher and developer guidance, and Google states that there are no additional requirements to appear in AI Overviews or AI Mode. Neither platform is holding closed a door that a larger budget would open, which is the part most firms get backwards. We worked through where a small budget goes first, and the order is not the one most firms are sold.

Frequently Asked Questions

Does domain authority matter for AI search?

Barely, on the available evidence. Surfer’s July 2026 analysis of roughly five million citations found correlations between authority metrics and citation frequency close enough to zero to be treated as noise.

Do I have to rank on page one to appear in an AI Overview?

No. Ahrefs found only 37.9 percent of Overview citations came from top-ten pages as of March 2026, which leaves the clear majority of cited pages sitting outside the first page of ordinary results entirely.

Why do AI answers name directories instead of law firms?

Directories took at least 51.8 percent of citations in InterCore’s 2026 study and were cited first in 77.8 percent of answers. They publish structured, comprehensive, frequently updated pages covering many firms at once.

How long before a small firm appears in AI search?

There is no sourced answer, and anyone quoting you one is estimating. What is documented is that no platform requires a minimum authority level for a site to be eligible.

Key Takeaways
1Legal directories took at least 51.8 percent of every citation in that dataset, because they publish structured, comprehensive pages covering many firms at once.
2Site authority barely predicts citation frequency, because correlations across roughly five million citation source URLs came back close enough to zero to be treated as noise.
3Only 37.9 percent of AI Overview citations come from top-ten pages, because these features issue multiple related searches rather than reading one result set.
4The opening for a small firm sits in the profiles already speaking for it, because 34.2 percent of the firms named in these answers had no page of their own among the cited sources.

Find Out What AI Already Says About Your Firm

Your firm already appears in AI answers, and most of what those answers know about you came from somebody else’s directory listing. We work out of Denver, and the first thing we do on any engagement is find out what the internet already says about a firm before we add a word to it. Call 877-486-8123 or contact us online and start there.

David Arato, JDs headshot

David Arato, JD is the founder of Lexicon Legal Content, an attorney-owned legal content marketing agency serving law firms since 2012. He has tracked how AI answer engines choose their sources across hundreds of law firm content programs, and is blunt about which advantages are real and which are simply being sold. David is a frequent contributor to Attorney at Law Magazine and Attorney at Work and a recurring guest on legal marketing podcasts.

Small Law Firm SEO: What to Do First and What to Skip

Posted on by David B. Arato

Search your city and your practice area in a clean browser window and count how many results above the fold your firm did not pay for. For most three-attorney firms the answer is zero, and the reason is that the advice they have been following was written for firms ten times their size.

Every guide ranking for this term hands the same five-part checklist to a solo estate planner and a forty-attorney injury firm. At Lexicon Legal Content we read those guides in the same week we read a client’s analytics, and the gap between them is usually a budget spent on scale the firm cannot maintain.

Q Short Answer

Finish the Google Business Profile first, then write the pages that answer the questions your callers already ask.

Most of what gets sold to a firm your size is scale you cannot maintain, and it competes for budget with the two things that compound.

Spend in This Order

Finish your Google Business Profile before anything else, then write the pages that answer the questions your callers already ask, then consider everything else.

Local ranking comes down to three factors in Google’s own documentation, which names relevance, distance, and prominence. The same page states plainly that “there’s no way to request or pay for a better local ranking on Google,” which disqualifies most of what gets sold as a local SEO package.

The order matters more than the individual tactics. A profile that is complete, correctly categorized, and collecting reviews will outperform a blog you abandon in month four, and it is the foundation any local AI search strategy gets built on top of.

What Not to Buy

The fastest way for a small firm to waste a marketing budget is to buy scale it cannot maintain.

Skip bulk city pages. Forty location pages you cannot keep current are forty pages that decay, and Google’s guidance on scaled content is not sympathetic to the intent behind them. Skip citation-count packages, because the local documentation above already tells you prominence is not purchasable by volume.

Head-term paid search is the second thing to leave alone, in a category where agencies publish cost-per-click figures in the hundreds of dollars for injury keywords. Those figures are self-reported by companies selling the service, and they are still enough to tell you the auction was not built for a two-thousand-dollar monthly budget.

The last thing to refuse is a retainer you cannot fund for eighteen months. Anchor the whole envelope to revenue instead of to a quote. Average marketing budgets sat at 7.8 percent of company revenue in the 2026 CMO Spend Survey Gartner released on 11 May 2026, up 1.3 points from the year before. That is the ceiling to reason from, and search is one line inside it rather than the whole of it.

What the Timeline and the Budget Actually Look Like

Treat every timeline and price you have been quoted as a rate card until someone shows you the dataset behind it.

Of the pages ranking for this search, exactly one links a budget figure to an external source. The rest publish monthly tiers and month-by-month projections that originate with the company selling the service. That does not make them wrong. It makes them unverifiable, which is a different problem and a worse one when you are the person writing the check.

Referral flow is why you can afford to be patient. Fifty-nine percent of solo and small firms named referrals as their top lead source in Clio’s 2025 Legal Trends report, released 8 May 2025. Search supplements that channel rather than replacing it, and budgeting a content program that compounds on that basis removes the pressure to buy fast.

Where AI Search Fits for a Firm This Size

AI Overviews appear on 23.6 percent of legal-category search results, and the question-shaped searches your clients actually type trigger them far more often than the head term you are chasing.

Those figures come from 146,122,391 desktop result pages Ahrefs collected in September 2025, where the published AI Overview trigger rates put question queries at 57.9 percent against 15.5 percent for non-question queries. The practical read is that “how long do I have to file a claim in Ohio” is a better target than “Columbus personal injury lawyer,” and it is cheaper to win.

None of this requires a separate program. “Optimizing for generative AI search is optimizing for the search experience, and thus still SEO,” reads Google’s guidance on generative AI search, updated 10 July 2026, which also states that no special schema markup is required to participate. We took apart whether smaller firms get cited against much larger competitors separately, using the citation data.

Frequently Asked Questions

How much should a small law firm spend on SEO?

Reason from revenue rather than from a flat figure. Gartner’s 2026 survey puts average marketing budgets at 7.8 percent of revenue across industries, and search is one line inside that number, not the whole of it.

How long before a small firm sees results?

Nobody ranking for this term sources their timeline. Local profile work can move within weeks, and content compounding is measured in quarters. Any specific month count you are quoted is an estimate, not a benchmark.

Should I do this myself or hire someone?

Do the Google Business Profile yourself, because it needs your details and your judgment. Hire out what requires sustained volume, and only if you can fund the engagement long enough for it to compound.

Can a three-attorney firm rank against a firm with forty?

On head terms, rarely. On the specific questions your clients actually ask, routinely, because those long-tail searches cost far less to win and the larger firm is usually not bothering to write for them at all.

Key Takeaways
1Every guide ranking for this term hands a three-attorney firm the same checklist it hands a forty-attorney firm, because the advice was written for scale you do not have.
2Local ranking runs on relevance, distance, and prominence, because Google weights those three factors and states plainly that none of them can be bought.
3Bulk city pages, citation-count packages, and head-term paid search all fail the same test, because each one costs more to maintain than a small firm can sustain.
4The spending order matters more than the individual tactics, because a complete profile and a handful of question-shaped pages compound while an abandoned blog decays.

Your Existing Pages Are the Cheapest Place to Start

Most firms this size are carrying pages nobody has opened in three years, and none of them show up as a line item anywhere. We have worked with more than 300 law firms across North America, which matters here mainly because it means we can tell you quickly which of your existing pages are worth keeping and which are decaying quietly. Call 877-486-8123 or contact us online and we will start with what you already have.

David Arato, JDs headshot

David Arato, JD is the founder of Lexicon Legal Content, an attorney-owned legal content marketing agency serving law firms since 2012. He has seen what a small firm’s marketing budget buys when nobody sequences the spending, and has built Lexicon’s programs to start with the work that compounds. David is a frequent contributor to Attorney at Law Magazine and Attorney at Work and a recurring guest on legal marketing podcasts.