Services AI Audit Roadmap FAQs About Reviews Blog Book Strategy Call
Article

Is Your Law Firm AI’s Choice? How ChatGPT Decides Which Attorney to Recommend

When a prospective client asks ChatGPT for a good estate planning attorney in Austin, the answer is not random. It is assembled from a handful of signals most firms have never checked.

Updated September 3, 2026 4 min read

Most law firm websites were built to persuade a person who has already arrived. Answer engines never arrive. They read, compare and decide before the prospect sees a single page. That changes what “being findable” means, and it is why firms with excellent Google rankings can be invisible in ChatGPT.

How an AI answer about attorneys is actually built

When someone asks ChatGPT, Perplexity or Google AI Mode who to hire, the engine does not run one search. It runs several, reads the top results from each, and assembles a shortlist from the names that appear most consistently with the most supporting detail. That process has three moving parts.

Signal one: machine readability

The first filter is blunt. Can the engine read your site at all? Firms lose here for boring reasons: attorney bios rendered as images, practice pages hidden behind tabs that need JavaScript, no Attorney or LegalService schema, and a robots file that blocks the crawlers AI platforms use.

Check this today. Open your robots.txt and search for GPTBot, PerplexityBot and Google-Extended. If any are disallowed, the two largest answer engines cannot read your site by your own instruction.

Signal two: authority signals

Answer engines weigh sources differently. In our audits of professional-services firms, the same few source types surface again and again when a firm gets recommended.

Source type Weight in attorney answers What to do
State bar and court listings Very high Match name, firm and practice area exactly to your site
Directory profiles (Avvo, Justia, Super Lawyers) High Complete every field, same wording everywhere
Google Business Profile High for local queries Primary category, services list, recent reviews
Local press and podcasts Medium, rising One quoted mention a quarter beats ten backlinks
Your own blog Medium when it answers questions Write the questions clients actually ask, in their words

Signal three: content positioning

The engine can only recommend you for the problems it has seen you solve. A firm whose site says “full-service litigation” is competing for nothing in particular. A firm whose site has a page titled “What happens if you die without a will in Texas” gets retrieved for the exact question a worried adult child types at 11pm.

Write for the question, not the keyword

Keyword pages were built for a ranking engine. Answer engines want a direct answer in the first paragraph, a named author with credentials, and a date that shows the page is current. That is the whole formula, and most firms have none of the three.

How to find out where your firm stands

  1. Write down the five questions a prospect would ask before hiring you. Use their words, not yours.
  2. Ask each question in ChatGPT, Perplexity and Google AI Mode. Note every firm named and every source cited.
  3. Count how often you appear, and read how you are described when you do.
  4. Open the cited sources. That list is your roadmap.

If you would rather have this done across 40 prompts and four platforms with a scored report, that is exactly what the no-cost AI visibility audit covers.

Key takeaways

Frequently asked questions

Does ranking well on Google mean ChatGPT will recommend my firm?

No, the two are related but not the same. ChatGPT and Perplexity read a wider set of sources and weight corroboration heavily. Firms ranking first on Google are often absent from AI answers because directory and bar listings do not match their site.

How long does it take to become visible in AI answers?

Machine-readability fixes show up within weeks. Authority and characterisation signals typically take three to six months to move, which is why the engagement is built as a six-month roadmap.

Do reviews affect AI recommendations?

Yes, mainly through characterisation. Engines carry the language of reviews into how they describe you. A pattern of “responsive” and “clear about fees” in reviews tends to appear in the answer itself.

Is this different for CPAs and wealth advisors?

The three signals are the same. The trusted source list changes: state CPA society listings and FINRA BrokerCheck play the role that bar listings play for attorneys.

Raja Sampathi

Founder, meanderingSapien · Austin, TX

Strategic consultant with 20+ years of enterprise experience, now focused on one problem: getting ChatGPT, Perplexity and Google AI to recommend professional-services firms by name. Every post draws on live visibility audits, not theory.

Self discovery

Start with Your No-Cost AI Visibility Audit

See exactly where you stand, and exactly where the opportunity is.

Get Your $3,000 Audit, No Cost