AI Search Visibility vs. Google Rankings: What 38,000 Law Firm Queries Reveal
A study of 38,000 AI searches across ChatGPT, Perplexity, Gemini, and Claude shows law firm AI visibility has almost nothing to do with Google page-one rankings.
Law firm marketing agency Forward Push ran 38,000 AI searches across ChatGPT, Perplexity, Gemini, and Claude between late April and early June 2026, covering 25 practice areas in 15 US cities. The core finding: ranking on page one of Google does not mean an AI tool will name your firm. With only 6% overlap between ChatGPT and Perplexity results for identical queries, the study argues that AI recommendations and Google rankings are effectively separate games, each driven by different sources and different logic.
What happened
| Data point | Figure |
|---|---|
| Total AI queries run | 44,411 |
| Clean answers after errors removed | 38,173 |
| Individual firm mentions captured | 96,988 |
| Unique law firms mentioned | More than 15,000 |
| Study period | Late April to early June 2026 |
| Practice areas covered | 25 |
| US cities covered | 15 (incl. Atlanta, Chicago, Dallas, Houston, LA, New York, San Francisco) |
| ChatGPT / Perplexity result overlap | ~6% |
| Gemini firm-naming rate | ~12% |
| ChatGPT / Claude firm-naming rate | ~99% |
Forward Push Law Firm Marketing published the study on July 28, 2026. Researchers asked all four AI tools the same questions a real potential client would type, then recorded who got named. The 6,238 responses lost to platform errors and quota limits were stripped before analysis.
Why it matters: four platforms, four different answers
Google returns ten links and lets the user decide. An AI tool returns a short list of names and stops. There is no page two. A firm that has spent years building Google page-one presence can be completely absent from the list an AI produces for the same search a minute later.
The study found that the top three firms in any given city and practice area captured about 34% of all mentions, while roughly 64 other firms split the rest, most named only once or twice. Visibility is extremely concentrated at the top.
Platform behavior varies sharply
| Platform | Firm-naming rate | Notes |
|---|---|---|
| ChatGPT | ~99% | Named a specific firm in nearly every answer |
| Claude | ~99% | Named a specific firm in nearly every answer |
| Perplexity | ~96% | Cites sources, making its reasoning visible |
| Gemini | ~12% | Usually defaulted to generic advice such as “check reviews” |
Because Perplexity shows its source citations, the researchers could read all of them. Across more than 30,000 citations, Super Lawyers appeared in 15.6% of all citations and Justia in 13.2%. Best Law Firms, Best Lawyers, BCG Attorney Search, and Chambers rounded out the top six sources. Ninety-six percent of Perplexity answers cited at least one legal directory. Only 68% cited the firm’s own website.
In other words, directories are writing a significant portion of the AI answer. A firm’s own site is a secondary source, not the primary one.
Our take
This study is a press release from a law firm marketing agency, so take the framing with appropriate skepticism. The methodology relies on self-reported data, and the sample is limited to one vertical. That said, the directional finding matches what we see across other sectors: AI tools pull heavily from structured, authoritative third-party sources (directories, review platforms, accreditation bodies) rather than crawling firm websites fresh.
The 6% overlap between ChatGPT and Perplexity results is the most practically useful number here. It confirms there is no single “AI result” to optimise for. Each platform builds its answer from a different corpus. As we covered in our look at recent Google Search volatility, organic visibility is already fragmented. AI answer engines add another layer of fragmentation on top.
For any business that cares about being found, the implication is the same whether you run a law firm or a plumbing company: your presence on structured third-party sources now matters as much as your own site. That applies directly to how we think about technical and content SEO for clients today.
Marc Apple, founder of Forward Push, put it plainly: “Being named by an AI tool is not the same as being the better lawyer. It means the machine found you somewhere it trusts.”
What to do about it
- Audit your directory presence: check whether your business is listed accurately on the top-cited directories for your sector (the equivalent of Super Lawyers and Justia in your industry).
- Test all four platforms: search for your business on ChatGPT, Perplexity, Gemini, and Claude using queries a real customer would type. Record who gets named.
- Prioritise the platforms that actually name firms: given Gemini’s 12% firm-naming rate, ChatGPT and Perplexity are where optimisation effort pays off first.
- Strengthen your own site as a corroborating source: Perplexity cited firm websites in 68% of answers. Make sure your site clearly states your location, practice areas, and credentials so it reinforces what directories say about you.
- Repeat the audit quarterly: AI training data and source weighting change without notice.
If you want help mapping how your business appears across AI answer engines alongside traditional search, talk to the Lumien team about an AI visibility audit.
Frequently asked questions
Does ranking on Google page one mean an AI tool will recommend your firm?
No. The Forward Push study found that Google rankings and AI recommendations are essentially separate. A firm can sit on page one of Google and be completely absent from what ChatGPT, Perplexity, or Claude returns for the same search.
Which AI platforms actually name specific law firms in their answers?
ChatGPT and Claude named a specific firm in about 99% of answers. Perplexity did so about 96% of the time. Gemini named a specific firm in only about 12% of answers, usually defaulting to generic advice instead.
What sources do AI tools use when recommending law firms?
Based on Perplexity's visible citations across more than 30,000 results, legal directories dominate. Super Lawyers appeared in 15.6% of all citations and Justia in 13.2%. Ninety-six percent of Perplexity answers cited at least one legal directory, while only 68% cited the firm's own website.
How different are AI search results across platforms for the same query?
Very different. ChatGPT and Perplexity, asked the same question in the same city and practice area, produced results that overlapped only about 6% of the time. There is no single AI result to win; each platform builds its answer from different sources.
