A few months ago, I typed this into ChatGPT:
"Who's the best personal injury lawyer in La Brea, California?"
ChatGPT gave me a real answer. A list. Names, firms, links. Confident, like it knew.
One problem: My friend, who is a PI lawyer in that exact area, wasn't on it. Not even close.
I opened another window and tried Google. My friend's site came up this time—but only after scrolling past a sea of paid ads and local directories. I checked who was #1 on Google Ads.
They weren't anywhere in ChatGPT's list.
That's when I realized: Google and ChatGPT don't agree on who's "the best."
That question stuck in my head.
Why does AI pick certain professionals and not others? Where does this information come from? Can it be influenced—ethically, strategically?
So I started testing.
I Asked AI Over 100 Times
I didn't just test lawyers. I expanded to med spas, dentists, financial planners, and consultants.
"Best med spa in Scottsdale?" "Top real estate agent in Denver?" "Who should I talk to for estate planning in Dallas?"
ChatGPT always gave an answer. So did Perplexity, Claude, Bing, and Google's new AI Overview. Sometimes they matched. Often they didn't.
I started tracking what they recommended and where they pulled the data from. I ran questions daily across multiple platforms. I played with phrasing to see how the answers changed.
Then I did what any lawyer with research brain would do: I built a system around it.
My research process:
•Analyzed 100+ AI answers across categories and platforms
•Tracked 82 different signals (source type, citations, format, style, tone)
•Repeated the process using different search engines to look for repeat patterns
We weren't trying to "game the system." We were trying to understand what the system pays attention to.
What I Learned
The way AI search engines decide who to recommend is not random. And it's not the same across tools.
Each one has its own structure:
•ChatGPT looks for authoritative, well-structured content
•Perplexity leans heavily on citations and sources
•Google's AI Overview mixes web summaries with directory logic
•Claude prioritizes recent, relevant expertise
But a few things were consistent:
✓ High-authority, well-structured content mattered ✓ Mentions in trustworthy third-party sources mattered more ✓ Earned credibility beats paid ads ✓ Question phrasing changes the results
The way you ask a question shapes the AI's answer. The way your business presents itself shapes whether you're in that answer.
If someone searches for a lawyer like you and you're not showing up… it's not an accident. It's a visibility issue.
Why This Matters Now
AI isn't replacing search. It's replacing decision-making.
When someone asks an AI, "Who should I work with?", they're skipping the scroll. They're asking a shortcut to trust.
The numbers are staggering:
•90% of Google searches now include AI overviews
•73% of professionals use AI for research
•1B+ daily ChatGPT queries
The models aren't neutral. They reflect the content they can see and verify. They build answers from public signals of trust, authority, and consistency.
AI doesn't give ten links. It gives one recommendation. Either you're in the answer—or you're invisible.
If you're a lawyer, service provider, or professional, that should feel urgent. Because this is happening right now, not in 6-18 months.
What I'm Building
I'm Leah Kabli, a UCLA Law graduate and practicing attorney who got curious about why some firms show up in AI answers and others don't.
After months of testing and tracking, I founded RottenPages to help law firms optimize for this new reality. We call it Generative Engine Optimization (GEO)—the science of making your firm the preferred source that AI tools reference and recommend.
Want to see where you stand right now?
We'll test your firm across ChatGPT, Claude & Perplexity using real legal queries to see if AI is citing your content, recommending your firm, or ignoring you entirely.
Ready for the complete analysis?
Our $997 Professional AI Visibility Audit includes:
•30-page comprehensive report showing your current AI visibility
•Competitive analysis of your top 5 local competitors
•Specific GEO implementation roadmap
•30-minute strategy call to review findings
•Analysis delivered in 5-7 business days
What you'll discover:
•What makes ChatGPT recommend some firms and not others
•How to test your own name or firm across all AI engines
•What kind of content AI trusts enough to cite
•How different AI search engines treat the same business
•How to earn visibility without gaming or gimmicks
Not theory. Not SEO tricks. Just what actually works in the AI era.
P.S. - I started this research because my lawyer friend was invisible to AI despite having excellent Google rankings. Now I help law firms bridge that gap. Don't let your competitors get there first.
Subscribe for weekly insights on legal AI marketing and GEO strategies.