Mine what you already have. Your Google reviews, your recorded intake calls, and every client email from the last quarter are marketing gold. Read them for the beliefs, values, and pain points, not the compliment. "They were great" tells you nothing. "I didn't think I stood a chance against a big company like Uber" tells you everything. Keep a tally of the core motivations rather than a folder of nice quotes. Then call ten recent clients and ask what almost stopped them from picking up the phone. Two clients naming the same fear is a pattern.

Name the one motivation and make it your segment. The environmentally conscious buyer Toyota found didn't fit an age bracket or a zip code, it fit a belief. Do the same with your clients. Write the worry down as a sentence in the client's own words, then pick the one you can answer better than any firm in your market. Here's the test: If your segmentation survey still leans into "35 to 55 in Tampa," you haven't struck gold yet.

Move the spend and the message to match. Toyota stopped paying for a campaign aimed at everyone and went where its buyers actually were, based on who they actually were. For you that means ad creative that leads with the beliefs and values instead of the category, placed where the people who share those values spend their time. And make sure the campaign strategy feeds into intake. What you say in the ad should mirror what you say on the phone.

Keep it simple enough that your owner will act on it. In their research, Yankelovich and Meer described one financial services firm that ran a statistically elegant segmentation and management refused to touch it, because nobody could follow how they built it. They redid it on plain criteria like the average net worth of clients and management adopted it immediately. If you can't explain your segment to the managing partner in a sentence, it dies in the deck.

"Injured? You need a lawyer."

"Don't talk to the insurance company alone."

"The adjuster's first offer is not the only offer."

"You may be entitled to compensation."

"Know your rights after a car accident."

"Free consultation. No fee unless we win."

"Board-certified in truck accident law."

"Forty cases tried to verdict in the last five years."

"The only firm in Tampa with a full-time nurse on staff."

"Every attorney speaks Spanish, not just intake."

Let your intake calls tell you which message to run. Pull last month's recordings, one practice area at a time, and listen for the question underneath the call. When callers want to know whether hiring a lawyer is worth it at all, run value. When they want to know why you instead of the firm they just hung up on, run differentiation.

In your hot areas, let the national firms sell the category. Morgan & Morgan spends enormous sums telling the country to call a lawyer, and that work lands in your market whether you fund it or not. Run the same message on a local budget and you're paying to repeat a national ad campaign. Put your money on the words only you can say. Read your homepage headline out loud and ask whether the firm across town could say it too. If it could, it isn't a message.

Say something a client can look up. Anybody can claim "best personal injury lawyer." Only a few firms in your market can say "board-certified in truck accident law," and a prospect can verify that one in about a minute. Injured clients believe what they can confirm.

In your cold areas, run value and be patient. Not every prospect chooses between firms, because some aren't in the market yet. Illness from something like Camp Lejeune water or a long-prescribed drug can surface years after the exposure, and few people connect the two on their own. That's why mass tort campaigns spend most of their budget on education. Other injuries are real but people write them off: a sore back after a fall that seemed minor at the time, or a claim the client figures isn't worth the hassle of hiring a lawyer over. Credentials answer none of that. Show them what a case looks like first, then argue that you should handle it.

Count the intake team before you blame the homepage. I'll pull up a firm's LinkedIn and count the intake specialists before we even talk. One is a bad sign. Two isn't much better. Then I ask one question: Of the cases you want, what percentage do you actually sign? Most owners guess 90-plus. In my experience it's closer to 60 or 70. Miss a call and take five minutes to ring back and that person has already reached somebody else. Nights and weekends with nobody on the phone, a hurt caller transferred three times by a phone tree, a voicemail box on a Friday afternoon? A new homepage fixes none of that.

Check whether you have enough traffic to judge the site at all. Four billboards and a referral network isn't a website problem. Neither is ranking at the bottom of page one, and here's why: If a case reaches you from down there, the firms in the top three already turned it down. That's the same reason conversion testing rarely tells a firm anything useful. Justin Croxton, a CRO guy I've worked with for years, will tell you most companies don't have the traffic to produce data worth acting on.

Fix before you replace. Cash is tight at every PI firm because of the fee cycle, so spend like it. Core Web Vitals, page speed, title tags, a WordPress cleanup — that work recovers most of what you're losing for a fraction of a rebuild. And if you have $20,000 sitting there, I'd put it into distribution nine times out of ten before I'd spend it tuning a site nobody's visiting yet.

If you do rebuild, protect what's already working. Own your code and your hosting. Some agencies will hand you the Figma files and keep the code in the back end, and then you don't really own the most valuable asset you have. On migration, 301-redirect every URL you move and leave your schema and your content silos intact, or you'll torch the rankings that feed you cases.

Nobody can ticket the car. Austin has logged 298 autonomous-vehicle incidents since July 2023, 231 of them Waymo's, and only eight became formal complaints. Texas requires a driver's signature on a citation, and there's no driver to sign. When an officer pursues one anyway, the summons goes to the vehicle's registered agent, which for Waymo is the company's law firm. Only on July 1 did a California law start letting police send noncompliance notices to the manufacturer.

The cars struggle around first responders. A Waymo blocked an ambulance from reaching a shooting in Austin in March. Another drove into an active construction zone near San Mateo in May while a highway patrol officer repeatedly ordered "Stop, Waymo" over a loudspeaker. In July, the head of the National Highway Traffic Safety Administration urged developers to improve how their cars handle emergency scenes.

The rules haven't caught up. No federal framework governs how these cars behave. States and cities write their own enforcement, one law at a time, while the fleets scale nationwide.

Johnson & Johnson Strikes a $5.5 Billion Talc Settlement Covering 76,000 Claims: Johnson & Johnson agreed to pay an estimated $5.5 billion to resolve about 76,000 ovarian-cancer lawsuits over its talc baby powder, a landmark deal that could end more than a decade of litigation, Reuters reported. The uncapped settlement could reach $7 billion or more depending on participation, and it pays out within 18 months rather than the decade-plus J&J's failed "Texas two-step" bankruptcies had proposed. It needs 95% claimant approval to take effect. Plaintiffs' lawyers called it a fair resolution. J&J, which still denies its talc causes cancer and calls the claims meritless, struck the deal after a string of courtroom wins, including a federal ruling last week casting doubt on whether plaintiffs can prove talc caused their cancer. The agreement covers existing claims only, not future suits.

A $604 Million Trucking Verdict Puts the Freight Broker on the Hook, Too: A Dallas jury returned a $604 million wrongful-death verdict against a Dallas trucking company, its driver, and freight broker C.H. Robinson over a March 2021 chain-reaction pileup in Mississippi that killed three people, the Dallas Morning News reported. Jurors apportioned fault at 45% to the driver, 32% to carrier Lupus Superior, and 23% to C.H. Robinson, finding the broker negligently hired the motor carrier and liable for the driver's conduct. Plaintiff firm Arnold & Itkin asked for $554 million. After three days of deliberation, the jury awarded $604 million, including $280 million to one victim's family. C.H. Robinson said it acted properly, noting the carrier held a satisfactory federal safety rating when hired, and vowed to appeal.

Florida's High Court Makes "Fraud on the Court" Dismissals Harder to Appeal: The Florida Supreme Court backed a more deferential appellate-review standard that lets trial courts throw out an entire case when a plaintiff or their lawyer lies about material facts, Bloomberg Law reported. The July 9 ruling grew out of a slip-and-fall against Publix, where the plaintiff lied about her recovery even though nobody disputed her surgery and injuries. The court’s majority rejected the stricter review the plaintiffs bar had sought. Plaintiff groups urged the court as amici to narrow the rule, warning that defendants face no equivalent "death sentence" for their own misconduct. Justice Jorge Labarga dissented, writing that the decision tilts against plaintiffs' constitutional right to have courts decide their cases on the merits.

Strict Liability Puts Taco Bell on the Hook in the Cyclospora Suits: After federal health officials linked the chain's lettuce to what lawyers call the largest cyclospora outbreak in U.S. history, three firms sued within hours, Reuters reported. The complaints don't claim Taco Bell caused the contamination, and they don't have to. Most states let a sickened consumer win on strict liability by showing only that the food was defective and made them ill, with no proof of negligence required. Damages cover medical bills, lost wages, and pain and suffering. Plaintiffs' lawyer Bill Marler said his prior cyclospora cases settled for $25,000 to more than $1 million, depending on severity. The current outbreak has hospitalized more than 140 people.

Wisconsin Residents Hit Microsoft With a Class Action Over Data-Center Noise: Residents of Sturtevant, Wisconsin filed a class action against Microsoft on July 1 over noise from its Fairwater data center in neighboring Mount Pleasant, the Milwaukee Journal Sentinel reported. The suit alleges the complex emits "excessive, consistent, and pervasive" noise from diesel generators and HVAC equipment, including chillers, cooling towers, and condenser fans, and that Microsoft failed to install adequate acoustic barriers. The plaintiffs seek damages. Microsoft says it aims to be a good neighbor and has taken steps to reduce the sound. Microsoft could build up to 15 more data centers in Mount Pleasant.

A Federal Case Exposes Fake Safety Certifications on Amazon E-Bikes: Amazon and the product-safety organization UL sued the Chinese sellers behind the Aipas e-bike brand in Seattle federal court, alleging they put UL's certification mark on e-bikes sold through Amazon and their own website without ever passing UL's safety tests, WIRED reported. A judge entered a permanent injunction on July 15 barring the sellers from using the mark, though they admitted no wrongdoing, and WIRED found Aipas ads on still carrying the UL logo afterward. The fake certs mask the exact risk buyers check for, lithium-ion battery fires, which a federal report linked to 45 mobility-device deaths from 2017 to 2024, 19 of them e-bikes. In June the Consumer Product Safety Commission proposed a mandatory federal safety standard for e-bike batteries, open for public comment through August.

You have three moves, and standing still isn't one. As capital consolidates PI, Chad lays out three paths: Build your own team by raising capital and rolling up firms, join a team someone else is building, or build a great firm and watch the market intelligently. He flags the first as brutal if it's your first rodeo, since running firms across different jurisdictions, cultures, and systems is no small thing. The one thing he won't call an option is standing still.

If you're going to join a team, the window is open now. Right now a firm has an open field of groups to join, Chad says. In 12 to 24 months, many of those teams will have formed and may already have a firm in your market. His test for picking one is refreshingly human: Look around the room and ask whether you'd actually want to work with these people for years.

Preserve the brand you bought. Where some roll-ups stamp one corporate name on every door, Chad keeps each firm's brand and culture intact. He points to how LVMH and P&G run portfolios of distinct brands. He's buying sustaining-success firms, not turnarounds, then supercharging the back office while leaving what made them great alone.

Get the incentives right, or you'll breed mice. Chad quotes Charlie Munger on incentives and tells the story of a college that paid students a dollar for every dead mouse — so students started breeding mice. It's his filter for every comp plan and partnership structure: Does it reward the outcome, or reward gaming the system?

As case costs climb, client service is the edge. More capital chasing cases means a higher cost to acquire each one. Chad's counter is retention. Clients who hire an attorney "tend to get three times more than the clients that are self-represented," he says, and a firm that delivers at scale keeps them, and their referrals, for life.

It shows you the band, not just the number. Every prediction ships with a 90% confidence interval and the cohort of comparable verdicts behind it, same case type, same county, same severity tier, so you can see when the data is thin and the answer is soft. Predict says its median accuracy runs 90 to 92% on higher-value auto and premises cases, drawn from more than 20,000 precedent cases.

It flags which cases clear your bar. Set your firm's acceptance threshold and Predict marks whether a case's predicted value clears it, so the team can concentrate on the matters worth the fight instead of guessing which ones to chase.

It drafts the demand letter. Every prediction can export a full demand letter in your firm's own format, PDF or Word, carrying the number, the band, the cohort, and the specific facts that moved the value.

It only works one side of the table. Predict is plaintiff-side only and says it will never sell to carriers or the defense, so the comparables you rely on aren't arming the other side of the table.