AI search readiness for personal injury law firms

Someone rear-ended on a Tuesday morning does not open eight firm sites. They open an assistant and type something closer to how they actually think: "a truck hit my car in Cook County, I think it was a delivery van, what kind of lawyer do I need and who is good?"

The engine reads whatever it can parse, writes a paragraph, and names two or three firms. Whether yours is one of them is decided before the phone rings.

The general version of this problem is on the law firms page: most firm sites are written to reassure a human who already arrived, and read thin to a machine. Personal injury has a sharper version of it, because personal injury queries are almost never about lawyers in general.

Personal injury queries are matter queries, not lawyer queries

Nobody asks an assistant for "a lawyer". They describe an event and a place.

Dog bite. Slip and fall in a grocery store. Rideshare passenger. Commercial truck. Nursing home. Construction site. Motorcycle. Wrongful death. Each of those is a different intake decision for you, a different venue, sometimes a different statute of limitations, and often a different answer about whether you would even take the case.

The engine has to answer at that resolution. And most personal injury sites give it one page called "Personal Injury" with a hero image, a phone number and a paragraph about fighting for you.

That page is a category label. The question was a matter type. The gap between the two is where firms disappear from the answer.

What "quotable" means to an engine here

An engine building a shortlist is not scoring your billboard. It is hunting for sentences it can quote and attribute to you.

For a personal injury firm those statements are specific, and almost all of them are already true. They are just not written anywhere a machine can find them.

• The matter types you actually take, named the way a client would say them, each with its own page rather than a bullet inside a menu.
• The jurisdictions you serve: states, counties, courts. Not a shaded map graphic.
• Bar admissions per attorney, with the state and the year, in text.
• Whether you handle a case through trial or refer it out, and at what stage.
• The fee structure in plain words. A contingency percentage, whether it shifts if the case goes to trial, and who fronts case costs.
• Consultation terms: free or not, phone or in person, languages available, and how fast someone answers after hours.
• Named attorneys with roles and credentials, not a wall of first names under headshots.

Read that list again. It is the same information a good intake call covers in the first four minutes. The gap is that "we say it on the phone" and "readable as text and structured data" are different states, and only one of them survives being read by a machine.

The service page structure guide covers the mechanics. The judgment call is which matter types deserve their own page, and the honest answer is: the ones you would actually sign today.

The referral moved

The old way assumed a click. Someone searched "car accident lawyer near me", saw a map pack and ten blue links, opened three sites and compared. You bought the ad, tuned the title tag and waited for the form fill.

The new way often resolves before the click. The engine reads what it can, describes which firms handle which matters, and the person calls the one or two names it produced.

SparkToro's zero-click study measured 58.5% of US Google searches in 2024 ending without a click. That is a whole-web figure, not a personal injury figure, so treat it as directional. The direction is what matters: when the answer is the destination, being readable to the machine writing it is the new front door.

So the job changed shape. It is not only "rank for car accident lawyer plus city". It is "be readable enough that the engine can use your page when it writes the shortlist". A site can do the first well and the second badly, and plenty of firms with big ad budgets do exactly that.

The specialist line, and where markup gets you into trouble

One thing worth flagging before you start adding structure, because structured data makes claims louder, not softer.

Advertising rules for lawyers are state rules, and they are not uniform. But the general shape is well known: under ABA Model Rule 7.2 (checked 2026-08-06), a lawyer may communicate that they practice in particular fields of law, but may not state or imply certification as a specialist unless certified by an approved organization.

That distinction lives in your copy, and it lives in your markup too. A schema block that describes a firm as a specialist in something it is not certified in is the same claim in a machine-readable wrapper.

Citedon does not review your pages for advertising compliance. It measures readability. Which means the sequence matters: get the copy right with your ethics counsel first, then make it readable. Doing it the other way round just publishes the problem in a format engines quote more easily.

What a scan actually shows on a personal injury site

Take the page you would most want an engine to use. Not your homepage. The practice-area page for the matter type that pays for the office.

Suppose the readout comes back 1 of 4: one engine returned the page as a source for a relevant question, three did not. The page ranks fine. Nothing looks broken.

The readout then shows why. No structured description of the firm as a legal service with an address and an area served. Matter types living only inside a dropdown menu. Counties named in an image. The contingency percentage in a PDF. Attorney bios that say "over 20 years of experience" without a single admission, year or bar number in text. An FAQ section built as a script-driven accordion that never renders as content.

To a machine that page reads as "a professional-looking law firm website". It does not read as "a firm that takes commercial vehicle collision cases in these three counties, staffed by these named attorneys admitted in this state, on a contingency basis at this rate".

Those are the gaps that keep a good firm out of the answer while a smaller one with a plainer site gets named. On WordPress, approve a fix and Citedon writes the missing structure in through the Yoast or Rank Math you already run, previewed first, added on top of what exists, reversible. The mechanism is in how AI engines decide what to recommend.

The damaging admission

Readability is not results, and we are not going to blur that.

Citedon makes no claim about verdicts, settlements, case outcomes or the quality of your representation. It does not evaluate whether your page is accurate or compliant. If a page overstates something, making it machine-readable helps an engine repeat the overstatement faster.

We also will not promise that any engine will name you. Legal is one of the categories where assistants hedge hardest, frequently answering with "consult a licensed attorney in your state" and a directory link instead of firm names. That is the engine behaving correctly. Readiness does not override it.

And if your caseload comes entirely from a lead vendor, a referral network or television, and you have no interest in a stranger finding you through an assistant, then an ongoing readiness subscription is not worth your money. We would rather say that than sell it to you.

Where to start

Scan the practice-area page for the matter type you most want more of, not your homepage. That is the page an engine has to work with when someone describes their accident.

The first scan is free on any site, no signup. Run the practice-area page a real claimant would land on and see which of the four engines returned it today. Run a free scan.