How to show up when a client asks AI about your firm

A woman in Coral Gables opens ChatGPT at eleven at night and types: "My father just passed, there is a house and a trust, and my brother is already arguing. Who should I call?" She gets three firm names back. If yours is not one of them, that is not bad luck. It is the sum of a few things the engine could not find, could not verify, or could not match. So what does it actually take to be one of the three?

A cheat sheet from Nader Alnajjar has been making the rounds on LinkedIn this month. It lays out six moves for getting a personal brand cited by AI, and the six are right. They are also written for founders selling themselves. A regulated practice carries a harder version of the problem, because every one of those moves runs into a rule. So here is the same list, rebuilt for firms that answer to the Bar, HIPAA, the SEC, or fair-housing law.

1. Be specific about what you do, inside the rules

"We help clients navigate complex legal matters" gives an AI engine nothing to work with. "We handle probate and trust litigation in Miami-Dade for families with estates over two million dollars" does. Specificity is how the engine connects your firm to the question in front of it. The same holds for a practice that says "personalized care" against one that says "concierge internal medicine for executives in Brickell, with a panel capped at three hundred patients."

Now, here is where the regulated version differs. Florida Bar Rule 4-7.14 restricts "specialist" and "expert" unless the lawyer is board certified, and the Texas rules read the same way. RIAs cannot imply a result. So the specificity cannot come from adjectives. It comes from the matter type, the geography, the kind of client, and the process. None of that is restricted, and all of it is what the engine is looking for.

2. Say it the same way everywhere

AI models reconcile a firm across every public source they can read: your website, Google Business, LinkedIn, and the directories your regulator runs. For a law firm that means the state bar directory. For a physician it means the state license lookup and Healthgrades. For an RIA it is the SEC's adviser search and FINRA BrokerCheck. For a brokerage it is the MLS and the DBPR record. Those regulator-run sources are high-trust to an engine, and most firms have never looked at what they say.

When the name, address, and description differ across those sources, the engine splits the firm into two or three weaker entities and cites none of them. The fix is one canonical description, written once, placed word for word on every surface. The repetition is not lazy. The repetition is the mechanism.

3. Get mentioned by others, and know which mentions carry a rule

AI does not just read what you say about yourself. It weighs what others say about you. Earned mentions, tags, co-authored pieces, and press all count. This is where a regulated firm has an edge it rarely uses, because professional mentions are exactly what these firms already collect: bar association panels, hospital affiliations, medical society talks, chamber events, peer referrals, and local press.

Client testimonials are the mention type that carries a rule. SEC Marketing Rule 206(4)-1 governs testimonials and endorsements for advisers, with disclosure of any compensation and conflicts. Florida Bar Rule 4-7.13 covers what a lawyer's testimonial may say. A patient testimonial needs a written HIPAA authorization and cannot reveal protected health information. The FTC Endorsement Guides at 16 CFR 255 apply to everyone. So the order is: pursue professional mentions first, because they carry authority with no disclosure load, and run client testimonials through counsel second.

4. Answer the question the client actually asks

AI surfaces content that responds directly to a query. Clients do not ask "what are your practice areas." They ask "what happens if I die without a will in Florida," "how much does concierge medicine cost in Miami," or "do I need an RIA or a broker." Write the page that answers each one plainly, in the first paragraph, and put the caveat the rule requires next to the answer rather than in front of it.

That last part matters. A disclaimer that blocks the answer makes the page useless to the engine and the reader. A disclaimer that sits beside a real answer makes the page both citable and defensible. Our earlier piece on what AEO playbooks miss covers how to structure those pages.

5. Build depth, not frequency

One detailed breakdown of how your firm handles a matter is worth more than thirty generic tips. AI rewards substance over volume. "How a Florida probate actually unfolds, step by step, with the timeline and how the fee is set" will be cited long after "five estate planning mistakes" has scrolled away.

For a regulated firm, depth is also the safer path. One thorough page gets reviewed once and stays compliant. Thirty shallow posts get reviewed thirty times, or not at all. Depth is a compliance strategy that happens to be a visibility strategy.

6. Think in prompts, not keywords

Nobody types "best estate attorney Coral Gables" into ChatGPT. They type the sentence at the top of this article. So the question to ask of your site is not "do we rank for the keyword" but "does any page of ours answer that prompt, and does the engine cite us when it does."

The practical tool is a prompt bank: ten or so questions per practice area, phrased the way a client asks them at eleven at night, checked one by one against your site and against what ChatGPT, Perplexity, and Google AI Overviews return. Most firms find that they have keyword pages for all ten and answer pages for none.

How Aday Interactive, Inc. runs the six

In our AI Visibility work, the six moves map onto three phases. Architect covers moves one, two, and six: the canonical description, the entity graph across your regulator's directories, and the prompt bank. Comply covers three and four: which mentions and answers are allowed in your vertical, and how the required language sits beside the answer instead of in front of it. Measure covers five and the result: which prompts now cite the firm by name, checked monthly across the engines.

The starting point is the free self-audit below. It shows what the engines are reading on your site today, before anyone decides what to fix.