The Rainmaker System

We Tested 3 Real Law Firms Against AI Search. All 3 Were Invisible.

In July 2026 we ran a live audit of the personal injury market in Marin County, California. We picked three established local firms, asked AI engines the exact questions an injured person would ask, and documented what came back.

None of the three firms appeared in a single answer.

These are not weak firms. One has served the county for decades. The audited group includes firms with Super Lawyers ratings and top Avvo scores. On paper, they are exactly who a Marin accident victim should be calling. The AI engines named other people.

Here is exactly what we asked, what the engines said, and why it happened. Everything below is reproducible; you can run the same queries yourself today.

The queries

We asked Perplexity two questions, phrased the way a real client phrases them.

Query one, broad intent:

"Who is the best personal injury lawyer in Marin County, California?"

Perplexity named one attorney: Gregory J. Brod of the Brod Law Firm. The citation behind the answer was his Justia directory page. Not his website, not his reviews, not his case results. A directory listing. None of the three audited firms were mentioned.

Query two, high intent:

"I was in a car accident in Novato. Which personal injury law firm should I call?"

This is the money question. Someone typing this has a case and is choosing a firm right now. Perplexity answered with three names: Arash Law, Moseley Collins, and Rush Injury Law. Two of those three operate outside Marin County. Large regional advertisers were outranking local firms on a query about a Novato car accident, in Novato's own backyard.

The scores

We scored each firm across the full stack that AI visibility depends on: presence in the sources engines cite, structured data, directory accuracy, and review signal.

Even the strongest of the three, the only one an engine actually named, did not rank on page one of Google for "personal injury lawyer Novato CA." Its home city.

Why this happened: the citation pool

AI engines do not evaluate law firms. They synthesize answers from a small pool of sources they already trust. For Marin personal injury, that pool is concrete and short: two Expertise.com roundups (best injury lawyers and best car accident lawyers in Marin), Yelp's algorithmic top 10 for Marin and San Rafael, the Super Lawyers San Rafael page, and the Justia and FindLaw directories.

The firms that dominate those specific lists (names like LNN Law, Sally Morin, Fleming Law, Delfino Green & Green) are the firms the engines can see. The root cause of invisibility was the same for all three audited firms: absence from the cited pool. The specifics differed:

• Rush Injury Law (51/100): present on both Expertise.com lists, but buried at 14 of 18 on one and 18 of 23 on the other. Its schema markup was malformed and pointed at a third-party website vendor's domain instead of the firm. No FAQ, attorney, or review structured data.
• Kuvara Law Firm (50/100): absent from both Expertise.com Marin roundups and from Super Lawyers San Rafael. It had genuinely useful FAQ content on its site, but with no FAQ schema, so machines could not use it. More than eight duplicate FindLaw pages fragmented its authority across listings.
• Anderson Franco Law (46/100): absent from both Expertise roundups, no Super Lawyers San Rafael listing, and its directory profiles were filed under San Francisco rather than San Rafael, so Marin-specific queries passed right over it. Its llms.txt returned a 404.

Notice what is not on that list: quality of lawyering, case results, years in practice. The engines never got far enough to consider any of it. If you are not in the sources being cited, your credentials are unread.

Run the test yourself

You do not have to take our word for any of this. Open Perplexity or ChatGPT and ask the two questions above, or adapt them to your own county and practice area:

1. "Who is the best [practice area] lawyer in [your county]?"
2. "I [had the problem] in [your city]. Which [practice area] law firm should I call?"
3. Look at who gets named, then click the citations and note which websites the answer was built from.

If your firm is named, congratulations, and check whether the sources naming you are stable. If it is not, the citations tell you exactly which pool you are missing from.

What this means if you run a firm

The uncomfortable takeaway from Marin is that invisibility is the default. Three real firms, decades of combined history, strong credentials, and the highest score among them was 51 out of 100. Not because they did anything wrong, but because nobody at any of them owns the specific, unglamorous work of being present in the sources AI engines cite: accurate directory profiles in the right city, placement on the roundup lists, working schema, consolidated reviews, an llms.txt that resolves.

The gap is fixable. That is the good news buried in this audit. Every deficiency we found has a known remedy, and we published a companion piece walking through all five: How to Get Your Law Firm Recommended by ChatGPT and Other AI Assistants.

It is also worth naming what these firms had already done right, because it makes the gap sharper. They had built real practices. They had earned credentials that the legal profession itself uses to signal quality. One had a genuinely useful FAQ page. The problem was never effort or merit. The problem was that the effort was pointed at humans, and the systems now answering client questions read a different set of signals entirely. A Super Lawyers rating that never makes it onto the Super Lawyers page for your city, or a review base scattered across duplicate listings, is credit the engines cannot spend.

The other takeaway is about timing. In every market we test, the cited pool is small and mostly unclaimed by the firms who deserve to be in it. The firms that move first get named while their competitors are still asking their web vendor what schema is.