How consistently does Google’s AI Overviews recommend the same law firms?

Ask Google’s AI overviews for a personal injury lawyer twice and you may get two different lists. Ask it 2,500 times across 50 cities, and the randomness resolves into something measurable: which firms are locked into the answers, which markets still reshuffle, and what the model reads before it decides. We ran a test to identify patterns in how Google’s AI Overviews recommended personal injury law firms when people searched for a personal injury lawyer in 50 different cities.

Key insights
  • Optimizing your website matters. AI reads law firm sites more than any other source when building its recommendation list. Two thirds of the 26,235 sources cited across our answers (65.8%) were law firm websites. Directories accounted for 20%, and everything else combined was under 14%.
  • AI is likely to research your market across the same 7 topics, no matter the city. Behind every answer the model runs its own searches, and they consistently fall into: ratings and rankings, how to choose a lawyer, settlements and success rates, practice specializations, fees and free consultations, firm size and reputation, and directories and referrals. Ratings and rankings alone carry 70% of the search volume, and “how to choose a personal injury lawyer in {city}” ran in every city we tested.
  • Some cities have clear winners that AI recommends every time. 40 firm-and-city pairs appeared in 100% of their market’s answers. The median city has 3 firms locked into 80% or more of its answers, though the spread is wide: Portland is the only market in the panel with no locked favorites at all, while Boston has five consuming most of every answer.

The test we ran: one question, 50 cities, 50 fresh answers each.

50Cities
2,500AI answers
9,308Searches observed
26,235Sources cited
1,535Firms recommended

AI answers are non-deterministic. One answer is an anecdote. Fifty answers are a measurement. We asked Gemini, the model family behind AI Mode and AI Overviews, the exact same question 50 separate times in each of the 50 largest US cities. Search grounding was on, so for every answer we captured three things:

  • the searches the model chose to run behind the scenes,
  • the sources it cited,
  • and the firms it recommended, in order.

Fifty passes per city gives every firm a recommendation rate with real error bars: appear in 40 of 50 answers and you are an 80% firm, give or take about 6 points. It also lets us define a locked firm (80% or higher, a favorite the model returns to almost every time) and an open slot (list capacity the favorites do not consume). Firm name variants were merged before any counting; 168 merges were reviewed by hand.

In many cities AI recommends the same law firms no matter how many times you ask.

How consistent is the AI? Consistent enough that most markets have a permanent cast. Forty-nine of our 50 cities have at least one locked firm. The median city has three. And 30 of the 50 cities have at least one firm that appeared in 100% of that market’s answers (40 firm-and-city pairs in all). In San Francisco, for example, one firm appeared in all 50 of the market’s answers, every single time in first position. (As a reliability check we also ran our home market, Salt Lake City, which sits outside this 50-city panel, with a different question phrasing on a different day; the same top five firms came back.)

30 of 50
cities already have a firm recommended in 100% of their 50 answers, a fully locked incumbent.
New York: no firm at 100%Los Angeles: has a firm recommended in 100% of its answersChicago: has a firm recommended in 100% of its answersHouston: no firm at 100%Phoenix: no firm at 100%Philadelphia: no firm at 100%San Antonio: no firm at 100%San Diego: has a firm recommended in 100% of its answersDallas: has a firm recommended in 100% of its answersFort Worth: no firm at 100%Jacksonville: has a firm recommended in 100% of its answersAustin: no firm at 100%San Jose: has a firm recommended in 100% of its answersCharlotte: no firm at 100%Columbus: has a firm recommended in 100% of its answersIndianapolis: no firm at 100%San Francisco: has a firm recommended in 100% of its answersSeattle: no firm at 100%Denver: no firm at 100%Nashville: has a firm recommended in 100% of its answersOklahoma City: has a firm recommended in 100% of its answersWashington, DC: no firm at 100%El Paso: has a firm recommended in 100% of its answersLas Vegas: has a firm recommended in 100% of its answersBoston: has a firm recommended in 100% of its answersDetroit: has a firm recommended in 100% of its answersLouisville: has a firm recommended in 100% of its answersPortland: no firm at 100%Memphis: no firm at 100%Baltimore: has a firm recommended in 100% of its answersMilwaukee: has a firm recommended in 100% of its answersAlbuquerque: no firm at 100%Fresno: has a firm recommended in 100% of its answersTucson: has a firm recommended in 100% of its answersSacramento: has a firm recommended in 100% of its answersAtlanta: has a firm recommended in 100% of its answersKansas City: no firm at 100%Mesa: no firm at 100%Raleigh: has a firm recommended in 100% of its answersColorado Springs: has a firm recommended in 100% of its answersMiami: has a firm recommended in 100% of its answersOmaha: has a firm recommended in 100% of its answersVirginia Beach: no firm at 100%Long Beach: has a firm recommended in 100% of its answersOakland: has a firm recommended in 100% of its answersMinneapolis: has a firm recommended in 100% of its answersBakersfield: has a firm recommended in 100% of its answersTulsa: no firm at 100%Tampa: no firm at 100%Aurora, Colorado: has a firm recommended in 100% of its answers
36.5
distinct firms recommended per city on average. The incumbent averages 49 of 50; the tail falls off fast.
The incumbent (rank 1) appears in 49 of 50 answers in the average cityThe market’s rank-2 firm appears in 46 of 50 answers, averaged across the 50 citiesThe market’s rank-3 firm appears in 42 of 50 answers, averaged across the 50 citiesThe market’s rank-4 firm appears in 37 of 50 answers, averaged across the 50 citiesThe market’s rank-5 firm appears in 31 of 50 answers, averaged across the 50 citiesThe market’s rank-6 firm appears in 26 of 50 answers, averaged across the 50 citiesThe market’s rank-7 firm appears in 20 of 50 answers, averaged across the 50 citiesThe market’s rank-8 firm appears in 17 of 50 answers, averaged across the 50 citiesThe market’s rank-9 firm appears in 14 of 50 answers, averaged across the 50 citiesThe market’s rank-10 firm appears in 11 of 50 answers, averaged across the 50 citiesThe market’s rank-11 firm appears in 10 of 50 answers, averaged across the 50 citiesThe market’s rank-12 firm appears in 9 of 50 answers, averaged across the 50 citiesThe market’s rank-13 firm appears in 7 of 50 answers, averaged across the 50 citiesThe market’s rank-14 firm appears in 6 of 50 answers, averaged across the 50 citiesThe market’s rank-15 firm appears in 5 of 50 answers, averaged across the 50 citiesThe market’s rank-16 firm appears in 5 of 50 answers, averaged across the 50 citiesThe market’s rank-17 firm appears in 4 of 50 answers, averaged across the 50 citiesThe market’s rank-18 firm appears in 4 of 50 answers, averaged across the 50 cities
1 city
has zero locked firms: Portland, the panel’s only genuinely open field.
New York: 3 locked firmsLos Angeles: 4 locked firmsChicago: 3 locked firmsHouston: 1 locked firmPhoenix: 4 locked firmsPhiladelphia: 2 locked firmsSan Antonio: 3 locked firmsSan Diego: 4 locked firmsDallas: 2 locked firmsFort Worth: 4 locked firmsJacksonville: 4 locked firmsAustin: 3 locked firmsSan Jose: 2 locked firmsCharlotte: 2 locked firmsColumbus: 5 locked firmsIndianapolis: 2 locked firmsSan Francisco: 4 locked firmsSeattle: 3 locked firmsDenver: 3 locked firmsNashville: 3 locked firmsOklahoma City: 4 locked firmsWashington, DC: 2 locked firmsEl Paso: 4 locked firmsLas Vegas: 3 locked firmsBoston: 5 locked firmsDetroit: 3 locked firmsLouisville: 4 locked firmsPortland: 0 locked firms (the open market)Memphis: 3 locked firmsBaltimore: 3 locked firmsMilwaukee: 4 locked firmsAlbuquerque: 1 locked firmFresno: 3 locked firmsTucson: 4 locked firmsSacramento: 3 locked firmsAtlanta: 2 locked firmsKansas City: 3 locked firmsMesa: 5 locked firmsRaleigh: 4 locked firmsColorado Springs: 5 locked firmsMiami: 4 locked firmsOmaha: 4 locked firmsVirginia Beach: 2 locked firmsLong Beach: 2 locked firmsOakland: 3 locked firmsMinneapolis: 3 locked firmsBakersfield: 4 locked firmsTulsa: 3 locked firmsTampa: 1 locked firmAurora, Colorado: 2 locked firms

In the square grids, each square is one of the 50 cities. Hover any square for the city behind it.

Here is the full spread. Most cities carry two to four locked firms; the extremes are where strategy changes:

0510151 city with 0 locked firms: Portland1none locked3 cities with 1 locked firm: Houston, Albuquerque, Tampa31 locked10 cities with 2 locked firms: Philadelphia, Dallas, San Jose, Charlotte, Indianapolis, Washington, DC, Atlanta, Virginia Beach, Long Beach, Aurora, Colorado102 locked17 cities with 3 locked firms: New York, Chicago, San Antonio, Austin, Seattle, Denver, Nashville, Las Vegas, Detroit, Memphis, Baltimore, Fresno, Sacramento, Kansas City, Oakland, Minneapolis, Tulsa173 locked15 cities with 4 locked firms: Los Angeles, Phoenix, San Diego, Fort Worth, Jacksonville, San Francisco, Oklahoma City, El Paso, Louisville, Milwaukee, Tucson, Raleigh, Miami, Omaha, Bakersfield154 locked4 cities with 5 locked firms: Columbus, Boston, Mesa, Colorado Springs45 lockedLOCKED FIRMS IN THE CITY (80%+ RECOMMENDATION RATE)

Number of locked firms per city across the 50-city panel. Only one city, Portland, has no locked favorites at all. Four cities have five. Hover a bar to see which cities are in it.

Counting locked firms only tells half the story, because list length varies too. A city with three locked firms and nine-firm answers still has six open slots; a city with three locked firms and six-firm answers is mostly decided. Plotting both puts every market on one map:

0123456789longer lists, less locked: easiest to entershort lists, locked favorites: hardestNew York: 5.7 firms per answer, 3 lockedLos Angeles: 6.9 firms per answer, 4 lockedChicago: 7.7 firms per answer, 3 lockedHouston: 6.9 firms per answer, 1 lockedPhoenix: 8.3 firms per answer, 4 lockedPhiladelphia: 9.4 firms per answer, 2 lockedPhiladelphiaSan Antonio: 8.1 firms per answer, 3 lockedSan Diego: 7.3 firms per answer, 4 lockedDallas: 7.9 firms per answer, 2 lockedFort Worth: 6.9 firms per answer, 4 lockedJacksonville: 7.0 firms per answer, 4 lockedAustin: 7.8 firms per answer, 3 lockedSan Jose: 7.9 firms per answer, 2 lockedCharlotte: 7.4 firms per answer, 2 lockedColumbus: 8.4 firms per answer, 5 lockedIndianapolis: 8.8 firms per answer, 2 lockedSan Francisco: 6.5 firms per answer, 4 lockedSeattle: 8.4 firms per answer, 3 lockedDenver: 8.0 firms per answer, 3 lockedNashville: 7.6 firms per answer, 3 lockedOklahoma City: 8.1 firms per answer, 4 lockedWashington, DC: 6.4 firms per answer, 2 lockedEl Paso: 7.2 firms per answer, 4 lockedLas Vegas: 9.1 firms per answer, 3 lockedBoston: 6.8 firms per answer, 5 lockedBostonDetroit: 7.1 firms per answer, 3 lockedLouisville: 8.0 firms per answer, 4 lockedPortland: 7.5 firms per answer, 0 lockedPortlandMemphis: 5.9 firms per answer, 3 lockedBaltimore: 7.0 firms per answer, 3 lockedMilwaukee: 7.7 firms per answer, 4 lockedAlbuquerque: 8.5 firms per answer, 1 lockedFresno: 7.2 firms per answer, 3 lockedTucson: 6.8 firms per answer, 4 lockedSacramento: 7.2 firms per answer, 3 lockedAtlanta: 7.9 firms per answer, 2 lockedKansas City: 6.5 firms per answer, 3 lockedMesa: 7.7 firms per answer, 5 lockedRaleigh: 7.6 firms per answer, 4 lockedColorado Springs: 8.7 firms per answer, 5 lockedColorado SpringsMiami: 7.4 firms per answer, 4 lockedOmaha: 7.1 firms per answer, 4 lockedVirginia Beach: 6.0 firms per answer, 2 lockedLong Beach: 7.1 firms per answer, 2 lockedOakland: 6.7 firms per answer, 3 lockedMinneapolis: 7.8 firms per answer, 3 lockedBakersfield: 7.0 firms per answer, 4 lockedTulsa: 7.7 firms per answer, 3 lockedTampa: 7.7 firms per answer, 1 lockedAurora, Colorado: 7.8 firms per answer, 2 lockedANSWER SLOTS CONSUMED BY LOCKED FIRMSFIRMS PER ANSWER
Most open third of markets Middle third Most locked third

Each dot is a city, colored by how locked its market is. Dashed lines mark the median city on each measure. Hover any dot for its numbers.

The corners of that chart are worth naming:

MarketOpen slots per answer Locked firms
Most open markets
Philadelphia7.62
Portland7.50
Albuquerque7.51
Most locked markets
Boston2.25
San Francisco2.94
New York2.93

Open slots are the average number of positions per answer not consumed by locked favorites.

Read this as a market-entry map. In Portland the AI has not made up its mind, and a firm is competing for seven genuinely available slots per answer. In Boston the realistic brief is displacing an incumbent with a 90%+ recommendation rate. Same model, same question, completely different competitive situation. Any AI visibility engagement that does not start by classifying the market is guessing.

There is a flip side to the locked firms. When a firm shows up in an AI answer at all, does it keep showing up? Usually not. We took every firm-and-city combination that appeared in at least one answer, 1,827 of them. That works out to 36.5 distinct firms recommended per city, a long tail of contenders competing under the incumbents. Then we counted how often each one came back across the 50 asks:

One-off: included in 1 of the city's 50 answers. 31.6% of firm-and-city pairsOne-off131.6%Rare: included in 2 to 5 of the city's 50 answers. 28.4% of firm-and-city pairsRare2 to 528.4%Occasional: included in 6 to 15 of the city's 50 answers. 18.5% of firm-and-city pairsOccasional6 to 1518.5%Frequent: included in 16 to 39 of the city's 50 answers. 13.1% of firm-and-city pairsFrequent16 to 3913.1%Locked in: included in 40 to 50 of the city's 50 answers. 8.4% of firm-and-city pairsLocked in40 to 508.4%

How often a recommended firm reappeared, counted by how many of the city’s 50 answers included it. Share of all 1,827 firm-and-city pairs that appeared at least once.

Six in ten of the firms that appear at all (60.0%) surfaced in five or fewer of the fifty answers. They are drive-bys: the model mentioned them once or twice and moved on. Only 8.4% of pairs are locked in. This is why a single favorable AI answer means very little. Seeing your firm once is closer to sampling luck than to a position, and knowing whether you are genuinely visible requires asking the question dozens of times, not once.

One more layer of recommendation math worth having on hand: national concentration is almost nonexistent while local concentration is extreme. The most-recommended firm in the entire panel holds 1.2% of the 18,713 slots, and the top five firms combined hold 3.7%. But zoom into a single city and the picture inverts: about one in twelve recommended firms is locked somewhere, and 40 of those firm-and-city pairs are at a perfect 100%. This market is not won nationally. It is won city by city, and in many cities, it has already been won.

Recommendation strength fades fast below a market’s top firms.

What “rank” means in this article

None of the rankings here are organic search rankings. When we say position, we mean where a firm was named inside a single AI answer: the first firm the answer recommends is position 1. When we say rank, we mean a firm’s standing in its city once all 50 answers are counted: the market’s most-recommended firm is rank 1. Both describe the AI’s recommendation list, not where anyone ranks in traditional search results.

Rank each city’s firms by recommendation rate, from most recommended on down, and compare the same rank across all 50 cities. The ladder is steep. The typical market leader appears in 100% of answers. The typical fifth firm appears in two thirds. The typical tenth firm is down to one answer in four, and by rank 20 a firm is appearing about once in every 25 asks:

0%25%50%75%100%Rank 1the leader100%range 78 to 100%Rank 294%range 66 to 100%Rank 386%range 54 to 98%Rank 566%range 28 to 84%Rank 1023%range 8 to 42%Rank 1510%range 4 to 20%Rank 204%range 2 to 12%

Recommendation rate by rank within a city, compared across all 50 cities. The tick is the median city at that rank; the shaded band is the full spread from the lowest city to the highest.

Two things follow from this shape. First, the top of the ladder is crowded with near-permanent firms: ranks one through three all sit above 85% in the typical market, which is why displacing an incumbent is a different job than getting mentioned. Second, mid-pack visibility is worth less than it sounds. Moving from a market’s 20th firm to its 10th makes you about six times as visible, and the biggest single jump on the whole ladder is the climb into the top five.

Wondering how your firm shows up in this data?

We can pull your recommendation rate, your rank, and who owns your market from the same panel.

Get Your Firm’s Report

Position wobbles too, and how much depends on where a firm sits. For every firm that appeared at least twice in a city, we measured how far its spot in the list drifted across its appearances. The locked leaders hold on tightly: centered near rank 3, drifting about 1.2 spots either way. Everyone below them floats in a wider band, roughly 1.5 to 1.7 spots of drift around a deeper average:

12345678910Locked in40 to 50 of 50rank 2.9drift ±1.2 spotsFrequent16 to 39 of 50rank 4.8drift ±1.7 spotsOccasional6 to 15 of 50rank 6.2drift ±1.7 spotsRare2 to 5 of 50rank 7.0drift ±1.5 spotsPOSITION IN THE ANSWER, AVERAGE AND TYPICAL DRIFT

Average list position and typical drift for firms in each recurrence tier, across all firm-and-city pairs with at least two appearances. Part of the leaders’ tightness is simple geometry, since a firm near rank 1 has nowhere higher to drift.

The practical read: below the locked tier, a firm does not own a spot in the list. It owns a band, and the goal of the work is to move the whole band up.

Rank and recommendation rate travel together, too. Firms that ever reach first position average a 65.3% recommendation rate across the whole panel; firms that never break into the top three average 7.5%. There is no meaningful consolation prize for fourth place. For a marketing program, the goal isn’t “get mentioned,” it’s get mentioned first.

Google’s fan-out queries for Personal Injury Law Firm Recommendations fall into the same seven topics.

Before writing an answer, the model runs three to five of its own Google searches. These are called fan-out queries, and they are the retrieval layer that decides visibility: whatever ranks for them is what the model reads. We captured 9,308 of them, 4,362 unique, and the pattern is remarkably uniform. Strip out the city names and just 29 templates cover 58% of all search volume. The single most universal query, the only one that ran in all 50 cities, is not a rankings query at all:

“how to choose a personal injury lawyer in {city}” is a search the model ran for itself while writing 707 of our 2,500 answers, and it ran in every city we tested. The model grounds its recommendations partly in educational content, not just best-of lists.

Clustering every query semantically produces seven topics, and their shares barely move between markets:

Fan-out topicShare of all searches Cities where it appears
Ratings and rankings (“best,” “top rated,” reviews)69.8%50 of 50
Selection guidance (“how to choose”)10.8%50 of 50
Settlements and success rates5.9%50 of 50
Practice specializations (car accidents, med-mal)4.4%50 of 50
Fees and free consultations4.2%50 of 50
Firm size and reputation2.8%49 of 50
Directories and referral lists2.2%46 of 50

The seven topics behind every fan-out query. Shares are occurrence-weighted: queries the model reaches for repeatedly count more.

Two details matter for content planning. First, 18% of all fan-out volume is year-stamped: the model appends “2026,” sometimes “August 2026,” to its own searches. Content with visible freshness answers the literal queries being run. Second, the topic proportions are nearly identical everywhere. The dominant topic’s spread across 50 cities is only 1.3 times what sampling noise alone would produce. Build content for these seven topics once and you have built it for every market you operate in.

The sources Google’s AI model references are mostly law firms.

Every answer cites its sources, 26,235 citations in our data. We classified all 1,387 cited websites:

Law firm websites
65.8%
Legal directories
20.3%
Award and ranking sites
4.4%
UGC and forums
3.6%
News media
2.8%
Off-topic placements
1.7%
Bar, government, and other
1.5%

Every citation behind the panel’s answers, classified into a 10-type source taxonomy, grouped here for readability.

The headline is the top row. The model’s reading list is mostly firms’ own websites. Whatever AI optimization means in legal, it starts on your own domain, because that is what gets read before the shortlist gets written.

The rest of the layer has sharp edges worth knowing:

  • Two directories carry the directory layer. Super Lawyers and Justia account for over three quarters of all directory citations. After those two, directory ROI falls off fast.
  • Award sites are read as authority. The third most-cited website in the entire study is an apply-to-be-recognized lawyer awards organization. It was cited more than every bar association and court combined. Official sources, bars and .gov domains, total 0.8% of citations.
  • Parasite SEO is measurable. 437 citations flow through 34 sites whose core topic has nothing to do with law: an arts blog, a restaurant site, lifestyle magazines, all hosting lawyer listicles. Small, but not zero. One cited “firm site” appears to be an outright template spam site.

The third-party legal sites doing the heaviest lifting with citations by Google’s AI Overviews

Set the firm sites aside and rank what’s left by citations, and the non-firm layer turns out to be a short list with very different shapes. Two national directories are cited in every single city. The awards site is cited almost as heavily but in only 15 cities. And two of the biggest remaining sources are not legal sites at all: YouTube is cited in 47 of 50 cities and Reddit in 34, which means video content and forum threads are part of the reading list nearly everywhere.

superlawyers.comlegal directory
2,54450 / 50 cities
justia.comlegal directory
1,53650 / 50 cities
eloa.orgawards site
74815 / 50 cities
bestlawfirms.comlegal directory
66734 / 50 cities
youtube.comvideo / UGC
65647 / 50 cities
attorneyatlawmagazine.comnews media
29027 / 50 cities
bestlawyers.comlegal directory
25627 / 50 cities
reddit.comforum / UGC
22634 / 50 cities
expertise.comawards site
15218 / 50 cities
thervo.comlegal directory
13219 / 50 cities
forbes.comnews media
13211 / 50 cities

The most-cited domains outside firms’ own websites, across all 2,500 answers. The right column shows total citations and how many of the 50 cities cite the domain at least once.

Underneath the national layer, each city has its own local legal sources that Google’s AI Overview uses

There is also a thinner layer the national view hides: domains the model cites meaningfully in exactly one city and nowhere else in the panel. This is where bar associations, city magazines, and metro news show up, the closest thing the model has to asking a local.

CitySource What it isCitations
San Franciscosfgate.comLocal news74
San Franciscosfbar.orgBar association44
Washington, DCwashingtonian.comCity magazine43
New Yorknycourts.govState courts33
Baltimorebaltimoremagazine.comCity magazine25
Washington, DCdcbar.orgBar association23
Dallasdmagazine.comCity magazine17
Oklahoma Cityokcexaminer.comLocal-branded listicle site17
Indianapolisindyexaminer.comLocal-branded listicle site16
Seattlekcba.orgBar association13
New Yorknycbar.orgBar association9
Portlandosbar.orgState bar8

Non-firm domains cited at least 8 times in exactly one panel city. “What it is” reflects our manual check of each site, not just the automated classification.

Two rows in that table deserve a closer look. The Oklahoma City and Indianapolis “news” sources are not newsrooms. Both are city-branded listicle sites, the same template running “5 best” roundups for dozens of unrelated service categories, gutter cleaning to wig shops, with generic bylines and no staff page. They pattern-match to local press well enough that the model reads them anyway. The local layer is real and earnable: bar association profiles and city magazine coverage genuinely get read. But it is also porous. The same one-city lens is where the off-topic placements from the list above tend to live, including a restaurant website cited 28 times in one market.

The source mix also connects back to consistency. Cities whose citations lean hardest on firm-owned sites are the locked ones (r = +0.41), and smaller cities lean more firm-site-heavy than large ones. Put those together: in smaller markets the model reads firm websites and picks winners, and the firm with the strongest site gets locked in. In bigger markets, aggregators dilute that advantage and the answers rotate more. A small-market firm largely controls its own AI destiny.

AI justifies its picks with fee promises and award badges, not review scores.

The answers don’t just list firms, they explain them. Scanning all 2,500 answer texts for the language the model uses to justify its picks shows which credibility signals it actually reaches for, and which conventional local-SEO signals it mostly ignores. The method is simple phrase matching against the full answer text, and one answer can mention several signals.

No fee unless you wincontingency language
99.7%
Free consultation offered
77.0%
Award or directory badgeSuper Lawyers, Best Lawyers, Avvo
76.0%
Case results / track record
47.5%
Years of experience cited
46.6%
Board or trial certified
21.8%
24/7 availability
21.4%
“Millions recovered” figures
17.2%
Firm size / resources
16.2%
Google reviews mentioned by name
6.0%
Bilingual / Spanish-speaking
4.4%
A specific star rating quoted
0.4%

Share of the 2,500 answers whose text contains each signal. Exact percentages shift a few points with the phrase list used; the ordering does not. The strictest reading of a quoted rating (“4.9 out of 5”) appears in under 1% of answers; even the loosest (“a 4.9-star rating”) reaches only about 9%.

Nearly every answer leans on the same three justifications. Contingency language is effectively universal at 99.7%, free consultations appear in 77% of answers, and award or directory badge names appear in 76%. That last one closes a loop from the section above: the badges the model cites come from the same directory and award sites it reads. Roughly half of answers cite years of experience or a results track record, and about one in five mentions board certification, around-the-clock availability, or a dollar recovery figure.

Now look at the bottom of the chart. Google reviews are named in only 6% of answers, and a specific star rating almost never appears, despite reviews being the trust signal local-SEO programs spend the most energy on. Review scores may still matter upstream, in deciding which firms make the shortlist at all, but the model does not repeat them when it argues for a firm. If AI is writing your firm’s pitch, it writes it from badges, fee structure, and results figures, not from your review count.

The biggest advertising law firms are everywhere, and locked in nowhere.

If AI answers simply mirrored brand awareness, the national TV advertisers would own this channel. They don’t. The brand with the widest footprint in our data appears in 25 of 50 cities at a 4% average recommendation rate. Another household name reaches 20 cities at 11%, leads in none of them, and typically appears sixth or seventh in the list. The average competitor in that brand’s own markets gets recommended twice as often as the brand does.

National TV brand A
25/50
The widest brand in the panel appears in half the country and averages a 4% recommendation rate where it shows up.
New York: not recommendedLos Angeles: recommended in 1 of 50 answers (2%)Chicago: not recommendedHouston: recommended in 1 of 50 answers (2%)Phoenix: not recommendedPhiladelphia: recommended in 1 of 50 answers (2%)San Antonio: recommended in 1 of 50 answers (2%)San Diego: not recommendedDallas: not recommendedFort Worth: recommended in 3 of 50 answers (6%)Jacksonville: not recommendedAustin: recommended in 2 of 50 answers (4%)San Jose: recommended in 1 of 50 answers (2%)Charlotte: recommended in 3 of 50 answers (6%)Columbus: not recommendedIndianapolis: not recommendedSan Francisco: not recommendedSeattle: recommended in 1 of 50 answers (2%)Denver: recommended in 2 of 50 answers (4%)Nashville: recommended in 1 of 50 answers (2%)Oklahoma City: recommended in 1 of 50 answers (2%)Washington, DC: recommended in 1 of 50 answers (2%)El Paso: not recommendedLas Vegas: not recommendedBoston: not recommendedDetroit: recommended in 4 of 50 answers (8%)Louisville: recommended in 1 of 50 answers (2%)Portland: not recommendedMemphis: recommended in 8 of 50 answers (16%)Baltimore: recommended in 1 of 50 answers (2%)Milwaukee: recommended in 1 of 50 answers (2%)Albuquerque: not recommendedFresno: not recommendedTucson: recommended in 1 of 50 answers (2%)Sacramento: not recommendedAtlanta: recommended in 1 of 50 answers (2%)Kansas City: not recommendedMesa: not recommendedRaleigh: not recommendedColorado Springs: not recommendedMiami: not recommendedOmaha: recommended in 1 of 50 answers (2%)Virginia Beach: not recommendedLong Beach: recommended in 8 of 50 answers (16%)Oakland: recommended in 3 of 50 answers (6%)Minneapolis: not recommendedBakersfield: recommended in 2 of 50 answers (4%)Tulsa: not recommendedTampa: recommended in 2 of 50 answers (4%)Aurora, Colorado: not recommended
National TV brand B
20/50
The household name: 11% average, the market leader nowhere, typically sixth or seventh in the list.
New York: recommended in 6 of 50 answers (12%)Los Angeles: recommended in 8 of 50 answers (16%)Chicago: recommended in 1 of 50 answers (2%)Houston: not recommendedPhoenix: not recommendedPhiladelphia: recommended in 3 of 50 answers (6%)San Antonio: not recommendedSan Diego: not recommendedDallas: not recommendedFort Worth: not recommendedJacksonville: recommended in 21 of 50 answers (42%)Austin: not recommendedSan Jose: not recommendedCharlotte: recommended in 3 of 50 answers (6%)Columbus: not recommendedIndianapolis: recommended in 2 of 50 answers (4%)San Francisco: not recommendedSeattle: recommended in 1 of 50 answers (2%)Denver: not recommendedNashville: recommended in 9 of 50 answers (18%)Oklahoma City: not recommendedWashington, DC: not recommendedEl Paso: not recommendedLas Vegas: not recommendedBoston: recommended in 1 of 50 answers (2%)Detroit: not recommendedLouisville: recommended in 9 of 50 answers (18%)Portland: recommended in 1 of 50 answers (2%)Memphis: recommended in 6 of 50 answers (12%)Baltimore: not recommendedMilwaukee: recommended in 1 of 50 answers (2%)Albuquerque: not recommendedFresno: recommended in 2 of 50 answers (4%)Tucson: not recommendedSacramento: not recommendedAtlanta: recommended in 15 of 50 answers (30%)Kansas City: not recommendedMesa: not recommendedRaleigh: recommended in 1 of 50 answers (2%)Colorado Springs: recommended in 1 of 50 answers (2%)Miami: recommended in 9 of 50 answers (18%)Omaha: not recommendedVirginia Beach: not recommendedLong Beach: not recommendedOakland: not recommendedMinneapolis: not recommendedBakersfield: not recommendedTulsa: not recommendedTampa: recommended in 6 of 50 answers (12%)Aurora, Colorado: not recommended
Strongest regional
9/50
The strongest multi-market firm: 50% average across nine cities, and the market leader in 1 of them.
New York: not recommendedLos Angeles: recommended in 50 of 50 answers (100%)Chicago: not recommendedHouston: not recommendedPhoenix: not recommendedPhiladelphia: not recommendedSan Antonio: not recommendedSan Diego: not recommendedDallas: not recommendedFort Worth: not recommendedJacksonville: not recommendedAustin: not recommendedSan Jose: recommended in 12 of 50 answers (24%)Charlotte: not recommendedColumbus: not recommendedIndianapolis: not recommendedSan Francisco: recommended in 6 of 50 answers (12%)Seattle: not recommendedDenver: not recommendedNashville: not recommendedOklahoma City: not recommendedWashington, DC: not recommendedEl Paso: not recommendedLas Vegas: not recommendedBoston: not recommendedDetroit: not recommendedLouisville: not recommendedPortland: not recommendedMemphis: not recommendedBaltimore: not recommendedMilwaukee: not recommendedAlbuquerque: not recommendedFresno: recommended in 47 of 50 answers (94%)Tucson: not recommendedSacramento: recommended in 22 of 50 answers (44%)Atlanta: not recommendedKansas City: not recommendedMesa: not recommendedRaleigh: not recommendedColorado Springs: not recommendedMiami: recommended in 2 of 50 answers (4%)Omaha: not recommendedVirginia Beach: not recommendedLong Beach: recommended in 41 of 50 answers (82%)Oakland: recommended in 9 of 50 answers (18%)Minneapolis: not recommendedBakersfield: recommended in 37 of 50 answers (74%)Tulsa: not recommendedTampa: not recommendedAurora, Colorado: not recommended
Not recommended Under 20% 20 to 50% 50 to 80% 80% and up

Each card’s grid is the same 50 cities in population order, colored by that firm’s recommendation rate in that city. Hover any square for the city and its rate.

The winning shape is the opposite: deep, not wide. The strongest multi-market performer in the panel appears in only 9 cities but averages a 50% recommendation rate in them, and the local firms that dominate a single market run at 90 to 100% while appearing nowhere else. That is what 90% of firms being single-city means in practice: this channel is won market by market, on grounding depth, not on national reach.

National TV brand A
New York: not recommended Los Angeles: recommended in 1 of 50 answers (2%) Chicago: not recommended Houston: recommended in 1 of 50 answers (2%) Phoenix: not recommended Philadelphia: recommended in 1 of 50 answers (2%) San Antonio: recommended in 1 of 50 answers (2%) San Diego: not recommended Dallas: not recommended Fort Worth: recommended in 3 of 50 answers (6%) Jacksonville: not recommended Austin: recommended in 2 of 50 answers (4%) San Jose: recommended in 1 of 50 answers (2%) Charlotte: recommended in 3 of 50 answers (6%) Columbus: not recommended Indianapolis: not recommended San Francisco: not recommended Seattle: recommended in 1 of 50 answers (2%) Denver: recommended in 2 of 50 answers (4%) Nashville: recommended in 1 of 50 answers (2%) Oklahoma City: recommended in 1 of 50 answers (2%) Washington, DC: recommended in 1 of 50 answers (2%) El Paso: not recommended Las Vegas: not recommended Boston: not recommended Detroit: recommended in 4 of 50 answers (8%) Louisville: recommended in 1 of 50 answers (2%) Portland: not recommended Memphis: recommended in 8 of 50 answers (16%) Baltimore: recommended in 1 of 50 answers (2%) Milwaukee: recommended in 1 of 50 answers (2%) Albuquerque: not recommended Fresno: not recommended Tucson: recommended in 1 of 50 answers (2%) Sacramento: not recommended Atlanta: recommended in 1 of 50 answers (2%) Kansas City: not recommended Mesa: not recommended Raleigh: not recommended Colorado Springs: not recommended Miami: not recommended Omaha: recommended in 1 of 50 answers (2%) Virginia Beach: not recommended Long Beach: recommended in 8 of 50 answers (16%) Oakland: recommended in 3 of 50 answers (6%) Minneapolis: not recommended Bakersfield: recommended in 2 of 50 answers (4%) Tulsa: not recommended Tampa: recommended in 2 of 50 answers (4%) Aurora, Colorado: not recommended
25 cities 4% avg rate
National TV brand B
New York: recommended in 6 of 50 answers (12%) Los Angeles: recommended in 8 of 50 answers (16%) Chicago: recommended in 1 of 50 answers (2%) Houston: not recommended Phoenix: not recommended Philadelphia: recommended in 3 of 50 answers (6%) San Antonio: not recommended San Diego: not recommended Dallas: not recommended Fort Worth: not recommended Jacksonville: recommended in 21 of 50 answers (42%) Austin: not recommended San Jose: not recommended Charlotte: recommended in 3 of 50 answers (6%) Columbus: not recommended Indianapolis: recommended in 2 of 50 answers (4%) San Francisco: not recommended Seattle: recommended in 1 of 50 answers (2%) Denver: not recommended Nashville: recommended in 9 of 50 answers (18%) Oklahoma City: not recommended Washington, DC: not recommended El Paso: not recommended Las Vegas: not recommended Boston: recommended in 1 of 50 answers (2%) Detroit: not recommended Louisville: recommended in 9 of 50 answers (18%) Portland: recommended in 1 of 50 answers (2%) Memphis: recommended in 6 of 50 answers (12%) Baltimore: not recommended Milwaukee: recommended in 1 of 50 answers (2%) Albuquerque: not recommended Fresno: recommended in 2 of 50 answers (4%) Tucson: not recommended Sacramento: not recommended Atlanta: recommended in 15 of 50 answers (30%) Kansas City: not recommended Mesa: not recommended Raleigh: recommended in 1 of 50 answers (2%) Colorado Springs: recommended in 1 of 50 answers (2%) Miami: recommended in 9 of 50 answers (18%) Omaha: not recommended Virginia Beach: not recommended Long Beach: not recommended Oakland: not recommended Minneapolis: not recommended Bakersfield: not recommended Tulsa: not recommended Tampa: recommended in 6 of 50 answers (12%) Aurora, Colorado: not recommended
20 cities 11% avg rate
The strongest regional
New York: not recommended Los Angeles: recommended in 50 of 50 answers (100%) Chicago: not recommended Houston: not recommended Phoenix: not recommended Philadelphia: not recommended San Antonio: not recommended San Diego: not recommended Dallas: not recommended Fort Worth: not recommended Jacksonville: not recommended Austin: not recommended San Jose: recommended in 12 of 50 answers (24%) Charlotte: not recommended Columbus: not recommended Indianapolis: not recommended San Francisco: recommended in 6 of 50 answers (12%) Seattle: not recommended Denver: not recommended Nashville: not recommended Oklahoma City: not recommended Washington, DC: not recommended El Paso: not recommended Las Vegas: not recommended Boston: not recommended Detroit: not recommended Louisville: not recommended Portland: not recommended Memphis: not recommended Baltimore: not recommended Milwaukee: not recommended Albuquerque: not recommended Fresno: recommended in 47 of 50 answers (94%) Tucson: not recommended Sacramento: recommended in 22 of 50 answers (44%) Atlanta: not recommended Kansas City: not recommended Mesa: not recommended Raleigh: not recommended Colorado Springs: not recommended Miami: recommended in 2 of 50 answers (4%) Omaha: not recommended Virginia Beach: not recommended Long Beach: recommended in 41 of 50 answers (82%) Oakland: recommended in 9 of 50 answers (18%) Minneapolis: not recommended Bakersfield: recommended in 37 of 50 answers (74%) Tulsa: not recommended Tampa: not recommended Aurora, Colorado: not recommended
9 cities 50% avg rate
A single-market leader
New York: not recommended Los Angeles: not recommended Chicago: not recommended Houston: not recommended Phoenix: not recommended Philadelphia: not recommended San Antonio: not recommended San Diego: not recommended Dallas: not recommended Fort Worth: not recommended Jacksonville: not recommended Austin: not recommended San Jose: not recommended Charlotte: not recommended Columbus: not recommended Indianapolis: not recommended San Francisco: not recommended Seattle: not recommended Denver: not recommended Nashville: not recommended Oklahoma City: recommended in 50 of 50 answers (100%) Washington, DC: not recommended El Paso: not recommended Las Vegas: not recommended Boston: not recommended Detroit: not recommended Louisville: not recommended Portland: not recommended Memphis: not recommended Baltimore: not recommended Milwaukee: not recommended Albuquerque: not recommended Fresno: not recommended Tucson: not recommended Sacramento: not recommended Atlanta: not recommended Kansas City: not recommended Mesa: not recommended Raleigh: not recommended Colorado Springs: not recommended Miami: not recommended Omaha: not recommended Virginia Beach: not recommended Long Beach: not recommended Oakland: not recommended Minneapolis: not recommended Bakersfield: not recommended Tulsa: not recommended Tampa: not recommended Aurora, Colorado: not recommended
1 city 100% rate
Not recommended Under 20% 20 to 50% 50 to 80% 80% and up

Each row is the same 50 cities in population order, one square per city. Darker means a larger share of that city’s 50 answers recommend the firm. The national advertising brands read as long rows of faint squares; the winning shapes are short and dark. Hover any square for the city and its rate.

How Law Firms Can Improve Your Recommendation Rate in Google AI Overviews

The answer to the title question: more consistent than most people assume, and the consistency is concentrated. A typical city’s AI answers have three firms that show up more than 80% of the time and a rotating cast filling the other four to five slots. Forty firm-and-city pairs are at 100%. A handful of markets are effectively decided, and a handful have no favorites at all.

At the national level the market looks wide open, with the leading firm holding 1.2% of all recommendation slots. That number is misleading. This channel is contested city by city, and inside a city the winners are stable, the mechanism is visible, and most of it sits within a firm’s control: the retrieval queries follow one national playbook with educational content at its center, two thirds of the grounding is firm websites, and the directory layer is two names.

So what do you do with all of this? The playbook falls straight out of the data:

  • Classify the market before you spend. An open market is a race for empty answer slots. A locked market is a displacement job against an incumbent the model returns to 90% of the time. They are different projects with different timelines, and any engagement that does not start by telling you which one you are in is guessing.
  • Build the site for the seven topics. Two thirds of what the model reads is firm websites, and its searches follow one national playbook: how to choose a lawyer, settlements and results, fees and free consultations, specializations. Answer those on your own domain, and keep visible dates on the content; nearly a fifth of the model’s searches ask for the current year.
  • Go deep on the two directories that matter, then go local. Super Lawyers and Justia carry over three quarters of the directory layer the model reads. Complete, current profiles there outweigh a long tail of listings everywhere else. Underneath that national layer, the model also reads each city’s own authorities: bar association profiles, city magazine coverage, and local news. Those placements get cited only in their home market, which is exactly the market a firm is trying to win, and they are earnable in ways a national directory ranking is not.
  • Write the evidence the model quotes. Answers justify their picks with contingency terms, free consultations, award badges, and recovery figures, almost never with review scores. If those proof points are not stated plainly on your pages, the model has nothing to repeat.
  • Play for first place, and measure it honestly. Firms that reach first position average a 65% recommendation rate; firms stuck below the top three average 7.5%. A single favorable answer is sampling luck. Ask the question dozens of times, track your rate, and judge the work by whether the whole band moves up.

Want to see how your firm performs in this data?

We hold the full panel: recommendation rates, list positions, market classes, and the sources the model read before deciding, for 1,535 firms across the 50 largest US cities. Tell us where to send your firm’s numbers, who owns your market, and what it would take to move.

Get your firm’s recommendation rate

Legal Query Fan-outs · Firm Profile

Law Firm A

How this firm shows up in Google AI recommendations

Footprint
9 of 50 cities
Present share
72% of answers
Locked markets
3
Best market
Denver · 96%