Why a ranking report isn't enough

A rank tracker tells you where a page sits in search results. An assistant answering "who handles this kind of case in my city" builds a sentence from what it can find and connect. If it can't settle who you are, where you work and what you do, it has little it can safely say about you. That's our read of how these tools behave, not a published rule from any vendor, which is why the test checks the answers directly instead of assuming them.

The five-minute test

Use a fresh session for each assistant: ChatGPT, Gemini and Microsoft Copilot are the obvious three. Log out or use a private window so personalization doesn't skew the answer. Ask these four questions, naming your firm and website exactly as you'd give them to a stranger:

  1. What is [firm name], and what does it do?
  2. Who is the lead attorney at [firm name]?
  3. Where does [firm name] practice?
  4. What practice areas does [firm name] handle?

Make a grid, assistants across and questions down. Mark each cell right, wrong or "didn't know." Twelve cells. The wrong ones are your to-do list, and a "wrong" that names a different firm with a similar name is worth screenshotting.

Run the market question too

Then ask the question a client would ask without your name in it, for example "who are the best [practice area] lawyers in [city]?" Note whether you appear and who does. Treat one run as a sample. Assistants can answer differently from one session or day to the next, so repeat it a few times before you draw a conclusion.

Reading the results

  • Right everywhere. Good. Repeat the test monthly so you notice changes.
  • "Didn't know." The assistant has no reliable source on you. See how to get your law firm cited by ChatGPT.
  • Wrong or mixed with another firm. Your public footprint disagrees with itself or is thin. Compare your site, your bar profile and your directory listings: are the firm name, city, attorney names and practice areas written the same way each time?
  • Right on one assistant, wrong on another. They draw on different sources. Copilot answers draw on Bing's index, which you can inspect in Bing Webmaster Tools; see the Bing AI Performance guide for law firms.

What this test can't tell you

It can't tell you whether a given answer will produce a client, and it's not a ranking. It's a snapshot of what an assistant returned on that day, in that session. A professional diagnostic repeats prompts across assistants and traces the sources behind each answer. AI Vortex offers one as the AI visibility diagnostic.

For a deeper look at one surface, read why most firms are invisible inside Copilot and Microsoft Copilot citations and legal content.

Sources

Questions and answers

How do I check if ChatGPT knows my law firm?

Open a fresh session, ask who the firm is, who the lead attorney is, where it practices and what it handles, and compare each answer to the facts. Repeat in Gemini and Copilot.

Why does ChatGPT describe my firm wrongly?

It's drawing on whatever sources it found, and if your public information is thin or inconsistent, or a similarly named firm exists, the answer can mix them up. The fix starts with making your own sources consistent.

How often should I run this test?

Monthly is a reasonable rhythm, and after you change your website or directory listings.

Does this test replace SEO tracking?

No. Search rankings and AI answers are different surfaces. Google says a page must be indexed and eligible to show in Search with a snippet to appear in AI Overviews or AI Mode, so the SEO basics still apply.

Can I pay someone to run this for me?

Yes. AI Vortex runs a diagnostic across the major assistants and traces the sources behind each answer. The details are on the diagnostic page.

Operational information, not legal advice. The test is our method and a one-day snapshot, not a ranking or a study. How assistants choose sources is not published in full.