The legal directories most frequently cited in AI search results are Super Lawyers, Avvo, and Justia. That’s according to the new WDW Legal Directories AI Visibility Study.
We Do Web staff searched 320 queries, each taking one of three forms (bare local, superiority, or conversational) and targeting one of three metro sizes (major, midsize, small) across 10 practice areas. We searched each query on four platforms (ChatGPT, Claude, Perplexity, and Google) and recorded the domains cited in each of the 1,280 AI answers.
More than half of the AI answers cited at least one legal directory, but the platforms we studied favored the directories to varying degrees, and some preferred (or didn’t) one or more directories over others.
The purpose of the study was threefold:
- Understand legal directories’ role in AI search answers
- Identify the legal directories most important to law firms’ AI visibility
- Discover other directories and domains that influence AI answers
We ran into a few surprises (e.g., Google AI Overviews rarely cited Avvo) and walked away with a few pieces of advice we’re happy to share with the legal community. But one takeaway stands out above all: claim your free directory listings. It just might boost your local visibility on the most popular AI platforms. What became most clear as we ran the study is that AI search loves legal directories.
TL;DR
- Searches with legal directory citations: 54.3%
- Platform most likely to cite legal directories: Perplexity
- Platform least likely to cite legal directories: Claude
- Most-cited legal directories: Super Lawyers, Avvo, and Justia
- Biggest surprise: Google AIOs rarely cited Avvo
- Practice area most likely to cite legal directories: DUI
- Metro size effect on legal directory citations: Google AIOs less & ChatGPT more likely to cite directories in small metros
- Non-legal directories with most visibility: Expertise.com, Yelp, and Reddit
The AI Platforms Most Likely to Cite Legal Directories
Our staff found that 695 of the 1,280 searches cited at least one legal directory, for a 54.3% legal directory citation rate across all four platforms.
Perplexity cited legal directories in nearly all its answers (293, 91.6%), followed by ChatGPT (230, 71.9%), Google AI Overviews (120, 37.5%), and Claude (52, 16.3%).
Takeaway: Law firms should audit their presence in legal directories (and that of each attorney on staff). With more than half of LLM and Google searches combined citing a legal directory, they represent enormous potential for visibility in AI search results.
The Legal Directories With the Most AI Search Citations
The most popular legal directories across all platforms were Super Lawyers (423 citations), Avvo (358 citations), and Justia (320 citations), with Lawyers.com (126 citations) and FindLaw (115 citations) a distant fourth and fifth, respectively.
The legal directories cited most overall were also cited most frequently on each platform, with some notable exceptions.
Super Lawyers was the most popular directory cited by Perplexity (58.44% of searches) and Google AI Overviews (25.31%) and tied with Avvo as the most frequently cited directory on Claude (5.31%). It was the second most cited directory in ChatGPT (43.13%), trailing only Avvo.
Avvo, meanwhile, was cited with relative frequency in ChatGPT (60.31%) and Perplexity (44.06%) and tied as the most cited legal directory in Claude (5.31%). The most surprising result was how infrequently Google AI Overviews cited Avvo: 2.19% of searches.
Other directories cited in >3% of searches were Lawyers.com, FindLaw, Martindale, Lawyer Legion, and Best Lawyers.
Takeaway: Law firms should claim free listings in each directory that offers them, especially those with the most AI search citations, prioritizing by visibility. Start with Avvo, Justia, Lawyers.com, FindLaw, and Martindale. Lawyer Legion and Law Firm Square are solid second-tier options.
A surprising find, however, was that Google AIOs rarely cited Avvo, making Justia the most popular free-listing legal directory in the most popular search engine.
Super Lawyers, meanwhile, offers the greatest visibility across the platforms studied, and attorneys recognized as Super Lawyers could gain an advantage in AI search visibility. Given the respect seemingly placed on the recognition, law firms should highlight attorneys recognized as Super Lawyers on their websites and other marketing.
Legal Directory Citations by Practice Area
Each query included one of ten practice area phrases:
- personal injury lawyer
- car accident lawyer
- slip and fall lawyer
- truck accident lawyer
- motorcycle accident lawyer
- divorce lawyer
- child custody lawyer
- dui lawyer
- criminal defense lawyer
- estate planning lawyer
DUI was more likely than other practice areas to trigger legal directory citations, controlling for platform (1.97x odds ratio, p = .004), and it remained significant even after correcting for the ten simultaneous comparisons. The effect was strongest on Justia and Avvo. No other practice area was significantly more likely than others to trigger legal directory citations.
Some directories, however, showed swings in citation rates by practice area. Motorcycle accident queries in Super Lawyers, for example, were well below the directory average, while child custody queries were well above the directory averages in Lawyers.com and FindLaw. Several other practice area queries fell well above or below the directory mean, as shown in the chart below. (The gray shaded area represents one standard deviation above and below the directory average.)
Takeaway: The per-directory differences between practice areas may interest law firms wishing to prioritize legal directories most likely to appear in AI search for their practice area. Law firms should consider not only the citation rate within each directory but each directory’s overall AI search citations (see Chart 2 above).
Legal Directory Citations by Metro Size
Each query contained one of eight city modifiers. We selected the eight cities so our searches spanned multiple metro sizes and varying degrees of competition.
- Major metros: Atlanta, Chicago, Houston
- Midsize: Nashville, Tampa, Charlotte
- Small: Reno, Savannah
We found no significant difference in legal directory citation rate by metro size across all platforms combined when controlling for platform (p = 0.749). However, Google AIOs and ChatGPT appeared less and more likely to cite legal directories in small metros, respectively.
We found both statistically significant after testing the correlations with logistic regression to calculate odds ratios.
- Small vs. major metros in Google AIOs: O.R. 0.38 (95% C.I. 0.20-0.72, p = 0.003)
- Small vs. major metros in ChatGPT: O.R. 3.16 (95% C.I. 1.54-6.49, p = 0.002)
Neither Google AIOs nor ChatGPT showed a significant difference in citation rate between midsize and major metros, and neither Claude nor Perplexity showed a significant difference by metro size, after applying Bonferroni corrections. A forest plot appears at the end of this post for those wishing to see the significance test results visually. See Chart 4.1.
Takeaway: Google’s lower odds of citing legal directories in smaller metros might suggest directories have less value for firms in those metros. But despite LLMs’ growing popularity, Google remains the most popular platform for search-like behavior, and even with lower odds of citing legal directories, a 22.5% citation rate represents a sizable volume of searches in small metros. So, law firms in smaller metros should pursue legal directory listings for Google AIO visibility.
Meanwhile, ChatGPT’s greater odds of citing legal directories in small metros suggest that if LLMs continue to eat into Google’s search market share, law firms in smaller metros could see greater visibility.
That said, our sample was quite small (just eight cities, two or three per metro size), and that limitation deserves scrutiny. A more expansive study of metro size’s effect on legal directory visibility in AI platforms is warranted.
Legal Directory Citations by Query Type
Each of the 320 queries fell into one of three categories:
- Bare Local: practice area+ lawyer + city (e.g., personal injury lawyer atlanta)
- Superiority: one of three superlatives (best, top-rated, best reviews) added to a bare local phrase (e.g., best personal injury lawyer atlanta)
- Conversational: one of two question bases (‘how do I find a good’ or ‘which should i hire’) around a bare local phrase (e.g., how do i find a good personal injury lawyer in atlanta, or which personal injury lawyer should i hire in atlanta)
Superiority and conversational queries were significantly more likely to cite a legal directory than searches with bare local queries when grouping all platforms. We also found significant differences across a couple of the platforms.
Correlations found to be statistically significant were:
- Superiority queries in all platforms combined: O.R. 3.51 (95% C.I. 2.34-5.26, p < 0.001)
- Conversational queries in all platforms combined: O.R. 2.50 (1.77-3.54, p < 0.001)
- Superiority queries in Google AIO: O.R. 5.62 (2.28-13.83, p < 0.001)
- Conversational queries in Google AIO: O.R. 11.82 (5.13-27.25, p < 0.001)
- Superiority queries in ChatGPT: O.R. 58.39 (7.73-440.85, p < 0.001)
Perplexity and Claude showed no significant difference in citation rate between query types. Forest plots appear at the end of this post for those wishing to view the significance test results visually. See Charts 5.1, 5.2, and 5.3.
Takeaway: Despite their relatively lower citation rates, bare local queries contribute far more to overall legal directory visibility. The bare local queries we studied, stripped of modifiers, averaged 245,400 monthly U.S. searches vs. 2,375 for superiority queries per Ahrefs. Conversational queries barely registered any search volume.
However, without LLM query data, we rely on Google search volume as a crude approximation. And because users are more likely to use longer queries in LLM searches than in traditional Google search, this may understate the prevalence of superiority and conversational queries in AI search and, in turn, their contribution to legal directory visibility.
Best Performing Non-Legal Directory Domains
Expertise.com was the fourth most-cited domain overall, trailing only Super Lawyers, Avvo, and Justia with 276 citations. It more than doubled the fifth most cited domain. However, it rarely appeared in our Google AI Overviews but was frequently present in ChatGPT citations.
Expertise.com is a “ranking site” that aims to list the top local businesses for a given location. Businesses cannot create a listing on their own, but some businesses, including law firms, can request a free review for inclusion.
Yelp was the fifth most-cited domain, with 132 citations in all. Law firms should claim and maintain their Yelp listing given a) how popular we found it to be in our study and b) ChatGPT’s recent deal with Yelp that allows the LLM to display Yelp-branded results in its answers. ChatGPT was the only platform that never cited Yelp, but we did not include the business listings in maps as citations in our study.
Reddit was the eighth most-cited domain in our study, though its visibility was only evident in Google and Perplexity. It was completely absent from citations in both ChatGPT and Claude. Law firms wishing to incorporate Reddit into their marketing strategy should be mindful of the forum’s strict rules regarding marketing.
Forbes was the 10th most-cited domain in our study, with 61 citations (54 of them in ChatGPT), and Attorney at Law Magazine was 11th with 55. Each publication offers Best Of lists that the AI platforms cited.
Conclusions
Claim Your Free Listings
AI search frequently relies on legal directories to respond to users whose search suggests they’re interested in hiring a lawyer. Law firms should include legal directories in their local SEO strategy, starting by claiming free listings in directories that offer them:
- Avvo
- Justia
- Lawyers.com
- FindLaw
- Martindale
- Lawyer Legion
- Law Firm Square
Our blog post on the legal directories provides greater insight into each.
Justia Is Your Best Free Listing
Google AI Overviews overwhelmingly preferred Super Lawyers and Justia, citing them in about a quarter and a fifth of searches, respectively. It rarely cited Avvo, however, and given that Google is still the most popular search engine by leaps and bounds, law firms prioritizing resources should focus efforts on Justia before Avvo.
Super Lawyers Reigns Supreme
Super Lawyers requires nomination and selection, but given its prominence as the most frequently cited legal directory across AI search platforms, earning recognition holds value beyond placing a badge on your website or a plaque in your waiting room.
Don’t Get Bogged Down in Practice Area or Metro Differences
Our analysis found swings in legal directory citation rates by practice area and metro size, but nothing in our results suggests a presence in the legal directories is a fruitless endeavor. Even at lower citation rates on some platforms, legal directories saw meaningful visibility in every practice area and every metro area we studied.
Google Reviews Are Still Primary
Most Google searches for queries suggesting interest in hiring a lawyer return a Google Map Pack. Yelp might have a deal with ChatGPT, but Google remains the most important search engine for law firms. If you’re going to focus your efforts on building reviews on any one platform, keep focusing on Google.
Another Benefit of Strong Google Reviews
One surprising result was Law Leaderboard’s performance, particularly in ChatGPT. Law Leaderboard isn’t a directory in the traditional sense. It analyzes Google reviews to score law firms. If you want your firm to be prominent in the directory, focus on generating as many positive Google reviews as you can and respond to each.
About the WDW Legal Directory AI Visibility Study
Queries Studied
We analyzed 320 transactional and commercial-intent queries that suggested the user wanted to hire a lawyer. When LLMs followed up to clarify the user’s intent, as was sometimes the case especially with bare local queries, the user responded in a way that suggested they were interested in hiring a lawyer. In doing so, we mimicked legal consumer behavior as closely as possible.
We made several efforts to diversify our queries across metro size, competition level, and practice type. We classified each query by target location/city, practice area, and query type.
The eight cities studied were:
- Major metros: Atlanta, Chicago, Houston
- Midsize: Nashville, Tampa, Charlotte
- Small: Reno, Savannah
The 10 practice area phrases were:
- personal injury lawyer
- car accident lawyer
- slip and fall lawyer
- truck accident lawyer
- motorcycle accident lawyer
- divorce lawyer
- child custody lawyer
- dui lawyer
- criminal defense lawyer
- estate planning lawyer
The three query types were:
- Bare Local: practice area+ lawyer + city (e.g., personal injury lawyer atlanta)
- Superiority: one of three superlatives (best, top-rated, best reviews) added to a bare local phrase (e.g., best personal injury lawyer atlanta)
- Conversational: one of two question bases (‘how do I find a good’ or ‘which should i hire’) around a bare local phrase (e.g., how do i find a good personal injury lawyer in atlanta, or which personal injury lawyer should i hire in atlanta)
We conducted 1,280 searches, each consisting of a specified query and platform.
Platforms Studied
We ran 320 searches apiece in Google, ChatGPT, Claude, and Perplexity.
Google searches were conducted in Incognito Mode and required users to change their location to the query’s target city using the gs location changer browser extension. A Google search was said to have triggered citations when it returned an AI Overview containing citations.
ChatGPT searches were conducted using the Temporary Chat feature. Staff members were not required to use a specific model but were instructed to record the model used in each search. All 320 searches used the GPT-5.6 Luna model.
Claude searches were conducted in Incognito. Likewise, members were not required to use a specific model but were instructed to record the model used in each search. Opus 5 was used in 47 searches, Opus 5.5 in 90 searches, and Sonnet 5 in 183 searches.
Claude was said to include citations in its answer when it searched the web and listed referenced web pages. In many cases, Claude searched places rather than the web, and may have relied on Google Places data; these searches were not said to have contained citations.
Perplexity searches were conducted in Incognito. All staff members used a free Perplexity account, which auto-selects the best model for the user’s query, so we considered the model for each search to be “Auto.”
Each platform at times referenced legal directories or other websites in its answer but did not list the directory or website among its citations; we did not count these references as citations in our study.
Data Collection
Six staff members participated in this study, which took place from September 9-24, 2026. Each staff member was randomly assigned an assortment of searches (defined by query + platform). In all, we ran 320 searches on each of the four platforms for a total of 1,280 searches.
Each staff member recorded the domains cited in the platform’s answer. Staff did not record citations to law firm websites individually but recorded the total number of unique law firm websites cited in each search.
Analysis
Our primary question was how frequently AI search platforms cite legal directories in their answers. We used simple probability and compared the percentages of searches that did and did not cite at least one legal directory. We then wanted to know which legal directories appear most often in each of the four platforms. We again relied on simple probability and compared total citations across and between platforms.
We also looked for patterns in legal directory citations by metro size, query type, and practice area. To do this, we compared the simple probability of a legal directory citation broken down by each variable. We used logistic regression to calculate odds ratios across all platforms combined for practice area, query type, and metro size analyses, controlling for platform. We also calculated odds ratios within each platform when analyzing metro size and query type. When testing whether each practice area was more or less likely than the group to cite legal directories, we compared each practice area with all others combined using logistic regression and controlled for platform.
When analyzing citation rates for practice areas within each directory, we calculated each directory’s mean and standard deviation across the ten practice areas, then flagged practice areas that fell one or more standard deviations above or below the mean; we did not test for statistical significance.
We considered correlations statistically significant if p < 0.05. We applied a Bonferroni correction to the 10 practice-area comparisons (p < 0.005) and the eight per-platform comparisons for each of the metro-size and query-type analyses (p < 0.006).
Limitations
Our study only recorded the domains cited when the AI answer included citations. We did not record brand mentions or legal directory or other domain mentions.
While our 1,280 searches overall provide a sizable sample, larger samples would increase power. Further, our variables – city/metro size, practice area, and query type – each contain a fraction of the overall sample, which invites scrutiny when testing correlation.
Tracking queries in LLMs is difficult, so we used Google search as a proxy for AI query volume. Search volume cited in the text is based on Ahrefs estimates of U.S. monthly search volume.
Our study did not count business listings in the map results included in some AI answers. This could undervalue certain listings, particularly Google Business Profile and Yelp listings.
Forest Plots
As promised, the following are forest plots representing the tests of statistical significance we ran when analyzing citations by metro size and query type.
The dark grey dot shows the odds ratio, and the teal dots show the lower and upper bounds of the 95% confidence interval. The vertical line at 1.00 represents no difference vs. the reference, which for metro size was major metros and for query type was bare local. Comparisons whose confidence intervals DO NOT cross the 1.0 threshold are statistically significant. Tests at the platform level must also remain significant after applying a Bonferroni correction, which applies a stricter standard to account for running multiple comparisons simultaneously. Conversational queries in Claude, for example, were not significant after the Bonferroni correction.
