👋 Good morning. Chris Dreyer here. A Florida firm we work with ran its first ChatGPT ads on Sept. 29. A week later, it had four signed cases. Wait until you see what it spent.
Also today: What makes someone an authority, the name everyone quotes around a specific issue? A content strategist who works with software founders says buyers check five things before they trust you. Most founders, he says, have only one. I'll show you how a PI firm can leverage this framework as part of its attention strategy.
And your LSA ratings could teach Google to send you fewer of the callers you want. We’ll talk about some things to watch out for in your campaigns. And if your caseload outgrows your team, I've got six fixes from operators who've been there.
Let's get into it.
📅 Lock In PIMCON 2027 at Early-Bird Pricing
I'd love to see you in the room next year. PIMCON 2027 is Oct. 3-5, back at The Phoenician in Scottsdale, Arizona, and early-bird tickets are on sale now. From one attendee this year:
"This is my second year at PIMCON, and each year it's been a great experience. We get to learn things about how other firms are approaching intake, marketing, and sales. The relationships that you build when you're at PIMCON is something that you continuously work on through the year. You don't see them once and that's it. They're typically relationships you build where you get to ask questions and learn from each other, and I think that's the really invaluable part about it." —Stark & Stark staffer
As a newsletter reader, you get an extra $150 off with code PIMNEWS. Early-bird pricing will be the lowest price to attend, so lock in your seat now.
📞 The Personal Injury Intake Optimization Playbook
Every dollar you spend on leads sits idle until an intake specialist turns one into a signed case. Move your conversion rate a single point and, depending on the size of your firm, you can add millions in revenue without buying one additional lead. No other function pays back that hard for a change that small.
I wrote down what the elite intake operations do differently, and I put all of it in one place. Fix intake and you raise the return on every other dollar you spend.
💡ONE BIG IDEA
How to Become the Lawyer Everyone Names
Logan Gott runs Gott Content, a content firm for software and AI founders. In a post on X on Oct. 11, he wrote about a pattern he keeps seeing. Many founders, he says, keep getting better at what they do, and wait for the market to notice.
Meanwhile, "someone with half their track record becomes the name everyone quotes, refers, and invites on stage."
He's worked with more than 30 founders, and he says the same thing every time:
"The gap almost never comes down to expertise."
Sound familiar? It should, because it's no different in PI. Many lawyers I talk to still lean on their legal credentials when they should master digital PR.
Earned citations, in other words.
This matters even more now, because AI answers work the same way.
When an AI model answers a question, most of what it cites comes from other people. In a May 2026 report, Muck Rack, a PR software company, analyzed more than 25 million links ChatGPT, Claude, and Gemini cited.
About 84% came from earned media, meaning sources a brand doesn't own or pay for, like news stories, research, Wikipedia, and social platforms. Journalism alone made up about 27%.
In other words, it isn't enough for your own website to say you're the best trucking lawyer around.
You need other people to say it, and keep saying it, in places you don't control.
Logan has a model for building that kind of reputation. He says buyers check five things before they trust you: expertise, visibility, association, voice, and proof.
And not only that, they ask themselves four questions before they buy or call.
Do I see you often enough to remember you?
What problem do I connect to your name?
Does this sound like an actual person I'd trust?
Has someone like me gotten a result from you?
Here's the wrinkle for PI. A firm has to answer those questions for more than one audience at once. There's the injured person deciding which lawyer to call. There's the reporter on deadline who wants to understand the spate of e-bike injuries. And now there's the AI model coughing up answers to queries from injured people.
Here's how I'd put Logan's framework to work for personal injury attorneys:
Get known for one problem, in one sentence. Logan tests whether someone can finish "that's the ___ person." For a firm, that's "the trucking lawyer." Kelley Simoneaux put it plainly on the Pod: "I am the spinal cord injury lawyer." Put that sentence in every attorney bio, LinkedIn headline, directory profile, and Google Business Profile. Use it with reporters, too. Chris Long, co-founder of the search agency Nectiv, wrote one description into eight profiles. Within a week, ChatGPT and Google's AI answers repeated it.
Give every post and every quote one job. Logan lists five: answer an objection, tell a client story, take a position on the news, show your process, or pick a side. At a firm, the objections come from intake, like "Will my rates go up if I file a claim?" Client stories need the client's written permission, a check against your state's advertising rules, and care with any medical details. As Logan puts it, "If a post doesn't do 1 of these 5 jobs, it's probably just filling the calendar."
Answer reporters while the story is new. Press is my addition to Logan's list. It grows out of one of his five jobs, taking a position on the news. When a crash, recall, or new law hits your market, reporters need a lawyer fast. The one already tied to that topic has the best shot at the call. In Muck Rack's data, citations peak in a story's first month and drop quickly through month six, so recent quotes count most.
Make your profile point somewhere. Logan reviews dozens of executive profiles a month and says 95% "are still set up like a college résumé." Someone reads your best post, clicks your name, and lands on a job history. Your LinkedIn page and firm bio should say who you help and show a result. Then give people a next step, like a guide to download or a number to call.
Judge it by calls and mentions. Logan's buyers "read quietly for months" before they reach out, so likes stay flat while trust builds. A firm has better numbers to watch. Count referral calls from other lawyers and reporter requests. Then check whether ChatGPT names your firm when you ask it about your one problem. Do that once a month.
The good news? None of this takes a massive content team. Logan's own routine runs about two hours a week. He says to run it for 90 days before you judge it, because most of the trust shows up after week six.
🔗 Logan Gott on X →
♟️STEAL THIS PLAYBOOK
How You Rate LSA Leads Changes the Leads You Get
Heads up! Google's LSA rating wants to know whether a call was a real, quality lead on its own merits, not whether you want the case.
A quality lead is a real person, in your service area, calling about an injury you handle. Case eligibility, whether the firm wants that case, is a separate question, and it's your intake team's job to make that call (no pun intended).
Mix those two things up, and Google may send you fewer of the calls you want.
That's our team's theory for what happened at one firm we work with. Its LSA signups fell to zero in a single week while it rated leads on case eligibility.
Once the firm switched to rating the quality of the lead itself, and doing it promptly, early signs pointed to a recovery.
So what does your team rate, the caller or the case?
I’d look into this soon, because that rating matters more than it used to. Google says your feedback helps it "send you more of the leads that are best suited to your business."
Since July 2024, its models also judge every lead and issue credits on their own. Google eliminated the old manual disputes.
Google wants that feedback on every lead. Ginny Marvin, Google's Ads Liaison, has encouraged advertisers and agencies "to provide feedback on all leads in their accounts."
Here's the problem…
When you mark a lead dissatisfied, Google's survey gives you a short list of reasons to pick from. They include: wrong service area, a service you don't offer, a caller not ready to hire, spam, a duplicate, and a solicitation like a job seeker.
None of the reasons Google names is a weak case.
Now take a real injured caller with a case you turned down. That call still fits one of Google's good-lead reasons, a lead "related to the services the business offers." As we read the form, marking it dissatisfied tells Google the lead itself was bad.
So let’s get to some fixes:
Rate every lead, every week, with one owner. Put one person in charge of the Leads page in your LSA account and block the same hour each week. They open each lead, click "Rate this lead," and submit. Google may add a bonus credit for leaving feedback.
Listen to the call before you rate it. The recording tells you whether the lead was real: who called, from where, and about what. A piece in the ABA's Law Practice Today gives the same advice: "listen to the call and review your internal notes before rating."
A real injury call in your area earns a "satisfied," signed or not. Google's satisfied reasons include a lead "related to the services the business offers" and one that "could convert into a booked customer or client soon." A crash victim in your service area fits the first one, even if intake turned the case down.
Save "dissatisfied" for the reasons on Google's list. Pick the one that fits each bad lead. A specific reason tells Google which leads to send less of. A blanket "dissatisfied" on a real injury call sends it the wrong signal.
Give case eligibility its own home. Log intake's decision, signed or turned down and why, in your intake software. That way the LSA rating stays a rating of the lead, and your team still keeps the case data.
This week, pull your last 30 days of rated LSA leads. Count the real injury callers marked dissatisfied, and fix the habit going forward.
That's an afternoon of work. A week of zero LSA signups costs a lot more.
What to Fix First When Your Caseload Outgrows Your Firm
When your caseload outgrows your team, more staff won't fix it by itself.
An operations lead at a young PI firm wrote to me with exactly this problem. Her firm carries 300 to 400 active cases, and two full-time case managers handle more than 100 each. She doesn't want to "simply hire more people and recreate the same chaos with a larger payroll."
The firm runs on Filevine, but the sheer volume leaves her almost no time to decide what the workflows, triggers, roles, and accountability should look like.
Where do you start if you run a growing PI firm?
How do you structure the team? What do you automate first? And how do you build all of it while the cases keep coming?
She isn't the only one asking.
So let’s flesh out some answers by turning to a few guests on the podcast who have already worked through similar challenges. And much of the advice comes down to this…
You don't rebuild the whole firm at once.
A better playbook:
Organize the team by case phase. Carol-Lynn Roman scrapped the one-manager-per-case model at Roman Austin Personal Injury Lawyers in 2022. Now a liability investigator sizes up coverage, and a teammate manages treatment. Then the file moves on to a demand writer, a negotiation team, and a settlement coordinator. Her Filevine setup adds the right teammate each time a case changes phase, so the handoffs don't drop. (Ep. 438)
Give operations its own leader. Michael Kelly credits his COO, Tim Paoli, with making everything else possible at Michael Kelly Injury Lawyers, now a 10-office firm. He tells owners to stop running every department themselves, because the CEO seat takes a different skill set than the lawyer seat. If a full-time operator is out of reach, Brett Schreiber's advice from September still holds: start with a fractional one. (Ep. 404)
Run each department on a few numbers. Kelly's department heads hold weekly meetings around the same six to eight KPIs, then present a monthly report to him and the COO. (He also says to make the Entrepreneurial Operating System, or EOS, your own, because a framework that feels too rigid won't get adopted.) Thaddeus Wendt of Feller & Wendt cut the pre-suit phase of a case from about 13 months to 8.3 months. He can only measure it because his team logs each step in Filevine. (Ep. 470)
Pull the repeat work into one team. Will Hammill and Jack Derrickson centralized case-management functions at Kenneth Nugent. They wanted to move cases faster, finish records and liens sooner, and push more files into suit. (Ep. 448)
Automate the routine calls first. EvenUp, the legal AI company, lists five starting points in its workflow automation guide: opening claims, confirming liability and coverage, chasing records and bills, treatment check-ins, and confirming balances. None of them needs legal judgment, EvenUp says. It advises law firms to measure the staff hours a task takes before you automate it, then add one job at a time.
Train one case type at a time. At Morgan & Morgan, Angie Flury's new intake agents train for six weeks before they take a live call. They learn one case type at a time, sitting beside a senior advisor. I'd copy that for case managers. Write down how your best person works one case type, teach it, then move to the next. (Ep. 447)
Pick the stage where files sit the longest, give it one owner, and track its time on desk every week. When that stage runs, build the next one.
📰 TOP OF THE NEWS
A Florida PI Firm Signed Four Cases in Its First Week of ChatGPT Ads
A Florida PI firm we work with signed four cases from ChatGPT ads in one week.
The firm launched its first ChatGPT ads on Sept. 29. Our paid team at Rankings runs the account, and the firm spent less than $115 on ChatGPT ads to sign those four cases.
Some fast facts…
All those ads ran statewide, at an average of $11.73 a click. Their 0.66% click-through rate sits just under the 0.73% U.S. average we covered two weeks ago.
Impressive, isn’t it? But a couple of thoughts I want to share with you:
One week is a small sample. Four cases on $115 works out to under $30 a case. Jessica Ford, our Director of Paid Digital, has seen cost per case run from $1,000 to $16,000 in other PI accounts.
The ads point to a page that explains the law. Clicks land on the firm's car accident page. It walks through Florida's no-fault insurance rules, the 14-day deadline to get PIP-covered medical treatment, and the state's 50% fault rule, then answers common questions. Jessica says ChatGPT approves that kind of educational page.
There’s another side to this. Jessica covers what ChatGPT approves when it comes to your landing pages in this week's Pod feature below.
🚀 QUICK HITS
Tesla Unveils a Thruster-Equipped Roadster Built for Human Drivers: Tesla unveiled its second-generation Roadster Oct. 15, The Wall Street Journal reported. With its compressed-air thrusters engaged, Tesla claims the car can go from a stop to 60 mph in about one second. CEO Elon Musk said the Roadster will eventually be the "only manually driven car that Tesla makes," with production starting in 2028. Asked how it might be street legal, Musk said, "Yeah, depends on your street, but it'll be a viable car."
Ohio's Litigation-Funding Law Takes Effect: Ohio's new rules for litigation funders took effect Oct. 6. Funders must register with the attorney general starting in early January, can't influence a client's case strategy or settlement, and can't do business with financiers not domiciled in the U.S. Within 14 days after a funded lawsuit settles or reaches a final order, the client's attorneys must disclose the agreement to the attorney general, who posts redacted copies online.
NHTSA Moves to Rewrite Federal Headlight Rules, Including Glare: The National Highway Traffic Safety Administration (NHTSA) announced Oct. 6 that it will update its vehicle-lighting standard, including how modern headlights create glare for other road users and how headlamp aim affects visibility. A committee of up to 18 stakeholders will negotiate the changes. Comments are due Nov. 9.
Florida's 2023 Bad-Faith Safe Harbor Applies to a 2018 Crash Claim: In an unpublished opinion Oct. 5, the Eleventh Circuit sided with Progressive in a bad-faith suit over a 2018 crash. The $7.49 million excess judgment came in April 2023, after Florida's tort reform law created a 90-day safe harbor. Progressive had tendered its policy limit and the property-damage demand within that window. A bad-faith claim accrues when the excess judgment is entered, the court said, so the new law applied.
New York City Sues a PI Firm, Alleging Fabricated Roadway Claims: The city announced Oct. 5 that it sued Asher & Associates in federal court, alleging the Manhattan firm blamed defective roadways for scooter, moped, and bicycle injuries that medical records tied to assaults and crashes. The suit cites 15 example claims and seeks treble damages under federal racketeering law. The claims are allegations, and the firm didn't comment to amNewYork.
Clio Buys Learned Hand, an AI Company Built for Judges: Clio acquired Learned Hand, whose AI workspace helps judges and clerks work through filings, conduct legal research, prepare bench memos, and draft orders, Artificial Lawyer reported Sept. 30. Courts across the U.S. use the tool, including the Superior Court of Los Angeles County. Clio CEO Jack Newton said that if lawyers move faster while courts keep the same capacity, "the bottleneck simply shifts to the courts." Clio also hired Casetext co-founder Pablo Arredondo and former Michigan Supreme Court Chief Justice Bridget Mary McCormack to lead its judiciary business.
🎙️ FROM THE POD
Jessica Ford on the Landing Pages ChatGPT Ads Will Approve
ChatGPT flags pushy injury landing pages and runs the helpful ones.
Jessica started testing ChatGPT ads for PI firms in late June, while OpenAI was still piloting legal ads case by case. In September, its ad policy caught up and now allows U.S. legal ads from lawyers licensed where the ad runs.
Jessica compares the channel to Local Services Ads when they first opened to legal: "absolute wild west, very few rules."
When Jessica came on the show, I wanted the granular version: what a firm can control, and which ads OpenAI shuts down.
Flags are less common now, but still a risk. When we recorded, OpenAI had flagged about 80% of her team's ads at some point after launch. Since then, Jessica says, OpenAI has become far friendlier to law firm ads, but ChatGPT ads still carry a higher risk of getting flagged. Ads that keep getting pulled can't spend, which is why early budgets came in low. Once the team kept ads live, PI accounts in major metros started spending $5,000 to $10,000 a week.
OpenAI approves educational landing pages. OpenAI's ad support confirmed to her team that the landing page affects approval. They flag pages that read as pushy or salesy, or lean on crash photos and emotional headlines. FAQ pages, blog posts, and a free ebook for people who were just in a car accident all ran well. She sees lead conversion rates of 5% or more, especially on simple opt-ins.
Contextual targeting triggers most disapprovals. It's the closest ChatGPT gets to keywords. Give it phrases like "people who are just in a car accident," Jessica says, and it flags the ads nine times out of 10. It works only as a hint, so her team removes it to keep ads running.
Narrow ads give you results you can trust. Each ad is a headline of up to 50 characters and a description of up to 100. Jessica runs at least four ads per ad group, each written for one case type in one market, like "Are you looking for a car accident lawyer in Denver?" So far, she says, the ads reach people before they know they have a case, in chats like "Somebody just hit me with a car, should I call a lawyer?"
Cost per case has ranged from $1,000 to $16,000. Both ends came from auto accounts, and she doesn't fully trust the attribution on the high one. Clicks have cost $3 to $64, and she's seen that whole swing in one account in a single week, and click-through rates sit between 0.1% and 0.5%. Before launch, she sets up the pixel and conversion tracking, including calls through CallRail.
"You actually want something where it's more educational, more helpful to the consumer to help teach them what they need to know. ChatGPT Ads loves that." —Jessica Ford
I told listeners I've never been this pushy in 400-plus episodes. If you waited on Local Services Ads, don't wait on this one. If you want to compare notes on a first test, reply to this email.
Here's our full conversation:
🤖 AI SEARCH TIP OF THE WEEK
Gemini can cite your page in one answer and drop it from the next.
Suganthan Mohanadasan, co-founder of the SEO firm Snippet Digital, built a free checker to see which pages Gemini keeps citing. When he asked Gemini the same question five times, it cited his own guide in three answers. An hour later, it cited the guide once.
His Gemini Grounding Gap Checker asks your question five times, then compares the facts Gemini keeps citing with what your page actually says.
The action this week: Run your main car accident page through the checker with the question it answers, like "What should I do after a car accident in Phoenix?" Start with the facts that come back in most runs. If your page touches one without stating it, add it in a sentence that makes sense on its own. Check each fact first, because Gemini repeats some outdated ones.
Then look at your robots.txt file. In his tests, publishers that blocked Google-Extended, the setting for Gemini, were cited in zero of 10 runs. Google says the setting doesn't affect search rankings, but allowing it also lets Google train Gemini on your content, so make that call deliberately.
🛠️ TOOL OF THE WEEK
Foundation AI Files and Routes Every Document Your Firm Receives
A misfiled medical bill can delay a demand or hide a lien.
Foundation AI, an Irvine, Calif., company founded in 2019, builds software for that problem. It pulls in what arrives by mail, fax, email, and portal download.
Then it works out what each document is and which case it belongs to, names it, files it in your case management system, and assigns the next step.
Foundation AI handles everything between a document arriving at your firm and the right person working on it.
Here's what stands out for a PI firm:
It catches attachments in staff inboxes. Its email capture watches the accounts you choose, pulls in case-related attachments from outside senders, and drops duplicates when the same document reaches several people.
It flags what it isn't sure about. Every case match and data field gets a confidence score. Anything below the firm's threshold goes to a person to confirm instead of a guess.
It turns documents into tasks. Your rules decide who gets the work, which deadlines get set, and when something escalates. Through supported integrations, it can also update treatment and economic-damages records. It connects to Filevine, Litify, CASEpeer, Clio, SmartAdvocate, and Smokeball.
Sweet James cut its document team from 28 people to six. The firm receives more than 80,000 PDF documents a month, nearly 400,000 pages, according to a Foundation AI case study. Founding partner Steve Mehr says the firm now spends "about 85% less time than before to process our incoming emails, medicals, and correspondence."
Foundation AI says more than 250 law firms use it, and its client list includes Morgan & Morgan and Dudley DeBosier. It's built for volume, and its own team handles setup and the ongoing process work. It lists no public price.
We haven't tested it. If your case managers spend hours a day on mail, book a demo and ask three things: the price per page or per document, how long setup takes, and whether it will sign a HIPAA business associate agreement before any client records leave your firm.
🔗 Foundation AI →
Disclaimer: Personal Injury Mastermind takes all reasonable steps to ensure accuracy in the materials we share, including articles, newsletters, and reports. These materials are intended for general informational purposes only and do not constitute legal advice. They may not reflect the most current laws or regulations. Always consult a qualified attorney for advice on a specific legal matter.
Thanks for reading. Quick ask…if you know someone who’d benefit from this content, please forward this to them. I’ll be back next week. - Chris
Received this newsletter from someone else? Subscribe below. Questions or want to sponsor this newsletter? Contact us at [email protected].