How to Rank in ChatGPT

1. Find the prompts you actually lose

Do not start with "AI SEO best practices". Start with the specific buyer prompts where your firm is invisible.

Send 20–40 buyer-style prompts into ChatGPT, Claude, Gemini, and Perplexity — 10 to 30 samples per prompt, because AI generation is stochastic and a single response is a coin flip. Record which competitors get named. Every named competitor is a page you can go read.

A useful prompt sounds like a buyer, not a category. Not "who are the best divorce lawyers in NYC" but "my spouse controls all the accounts and I'm trying to leave — who should I talk to in New York?" The first is a directory lookup; the second is what AI assistants are actually asked.

Every prompt where you are named zero times becomes a target. That is your gap list.

2. Rewrite the losing pages in answer-shape

An AI assistant retrieves at the sentence level, not the page level. Which means it needs a sentence to reuse. "Complex matrimonial matters" is not that sentence. "When one spouse controls the finances and the other is trying to leave safely" is.

Three rewrites reliably move citation rate:

• Category labels → buyer situations. "Family Law" is a Google label. Buyers don't type it. "Divorce with hidden assets" or "custody with an out-of-state parent" is what they type.
• Credentials → decisions. "35 years of experience in matrimonial law" is a credential. "We handle contested divorces where a settlement has already been rejected" is a decision-relevant sentence. Assistants reuse the second.
• Boilerplate → specific detail. Verifiable, situational, boring-but-precise language beats brand-voice adjectives. "Same-day emergency filings for domestic-violence protective orders" wins over "compassionate legal counsel in your time of need."

Every page targeting a buyer prompt needs at least one paragraph an assistant can quote verbatim without embarrassment.

3. Publish answer blocks and FAQPage schema

ChatGPT and Perplexity preferentially retrieve from FAQ-shaped content because a Q/A pair is a self-contained answerable unit. Every landing page targeting a buyer prompt should carry three to six FAQs, each satisfying:

1. The question sounds like a real buyer prompt.
2. The answer is a single, quotable paragraph — not a link dump.
3. The pair is wrapped in FAQPage JSON-LD.

Schema alone does not rank you. Schema plus the right prose does. If the visible answer is boilerplate, the JSON-LD wrapper doesn't save it.

4. Add citation-worthy proof

An AI assistant that names a firm is making a small commitment. It will prefer firms with something to point to. Three kinds of proof surface reliably in AI answers:

• Verifiable numbers. Not "extensive experience" — "612 cases resolved in the last three years." Not "high success rate" — "94% custody rulings favorable to primary caregiver in 2025."
• Named work. Landmark cases, published opinions, media appearances, professional recognitions — with sources.
• Live disclosures. Fee structures, response times, intake protocols, and other operational specifics that other firms bury.

Vague brand voice does not beat verifiable specifics in an AI answer. Ever.

5. Re-run and prove the delta

The final move separates AI ranking work from theatre. After publishing revised content, wait 2–6 weeks and re-run the exact same prompt set against the exact same assistants, at the exact same sample size. If citation rate moved above the noise floor, keep going. If it didn't, the rewrite missed and needs another pass.

Anyone selling AI ranking work who cannot show you the before-and-after prompt set with matched sample sizes is guessing. That is why Viclaro publishes methodology-versioned Atlas snapshots and why the audit product ships with a re-run built in.