Most law firm websites in 2026 fall into one of three buckets on schema markup: zero schema at all, generic Organization schema copied from a WordPress plugin, or the deprecated Attorney schema that Google stopped honoring. None of the three help an AI engine decide whether to cite your firm. The right stack is LegalService for the firm, Person for each attorney, FAQPage on practice-area pages, and Review aggregated from your real review sources. Done correctly, this is a 30-minute change for a small firm and a 4-hour change for a large one. Done incorrectly or skipped entirely, it is the single biggest reason your firm does not get named when a buyer asks ChatGPT for a recommendation.
This post is the technical map. What schema types to use, which ones Google quietly killed, what the JSON-LD actually looks like, and the order to ship them.
What schema markup actually does for a law firm in 2026
Schema markup is structured data that sits in the head or body of your HTML and tells search engines and AI models what each piece of content represents. Without schema, a model parsing your bio page sees a wall of paragraphs and has to guess that “Sarah Patel” is a person, “Patel Injury Law” is the employer, and “California Bar #294612” is a license number. With schema, those facts are tagged: Person, worksFor, hasCredential. The model does not have to guess.
That mattered before AI search. It matters more now. ChatGPT, Perplexity, Google AI Mode, and Claude all use entity recognition to decide which firms to cite when a buyer asks “who is the best probate attorney in Phoenix.” Sites with structured schema get cited 3.2 times more often than sites without it, according to citation studies published across 2025 and 2026. FAQPage schema specifically shows up in 67 percent of AI answers for relevant questions. The numbers compound: schema is not a tiebreaker, it is a gating signal.
The other thing schema does in 2026 is anchor your firm to the entity graph. Google, Bing, and the AI engines all maintain knowledge graphs that connect “Patel Injury Law” to its address, attorneys, services, and reviews. If your site does not feed that graph through schema, the engines pull data from third-party directories, often with stale phone numbers, wrong office hours, or missing attorneys. Schema is how you take ownership of your firm’s entity record from the source.
The schema types every law firm needs (and the one that is dead)
Here is the working stack for 2026.
LegalService is the primary schema type for the firm itself. It is a subtype of LocalBusiness specifically built for legal service providers, and it accepts every property a firm needs: name, address, phone, opening hours, area served, accepted payment, and the list of services offered. LegalService is what gets your firm eligible for local pack listings, knowledge panel features, and AI citations as a “law firm in [city].” Use this on your homepage and your Contact or About page.
Organization sits above LegalService in the hierarchy. For a law firm, you typically pick one or the other on a given page, not both. The exception is multi-office firms, where Organization describes the parent firm and each office uses LegalService nested under it.
Person is the schema type for each attorney bio page. This is critical and often missed. Person carries the attorney’s name, job title (jobTitle: “Partner” or “Senior Associate”), employer (worksFor pointing back to your LegalService entity), education (alumniOf), credentials (hasCredential for bar admissions), areas of practice (knowsAbout), and contact info. The model uses Person schema to confirm which lawyer at your firm handles which type of case. A bio page with no Person schema is invisible to AI engines as a routable entity.
FAQPage is the schema type for question-and-answer blocks on practice-area pages. If your DUI page has a “Common questions about DUI defense” section, wrap it in FAQPage schema. Each question becomes a Question entity with an Answer entity attached. AI engines pull these answer blocks directly into responses, which is the single highest-leverage AEO move available.
Review and AggregateRating are the schema types that surface your reviews to AI engines. AggregateRating goes on the LegalService entity and rolls up to a single rating value with a review count. Individual Review entities carry star rating, reviewer name, and the review body. The trap here: do not invent reviews. Schema review fraud is a fast track to a manual penalty. Pull from Google Business Profile or your review platform of record.
Now the dead one. The Attorney schema type was deprecated in 2024 and Google no longer honors it for rich results. It still validates as legitimate schema.org markup, but it triggers no enhancements and provides no citation lift. If your developer or a legacy SEO plugin is still pushing Attorney schema, swap it out. Use Person schema for the attorney and connect that Person to your LegalService entity through the worksFor property. That is the post-deprecation pattern.
What the JSON-LD actually looks like
Here is the LegalService block for a single-office personal injury firm. Drop this into a script tag with type=“application/ld+json” in the head of your homepage.
{
"@context": "https://schema.org",
"@type": "LegalService",
"@id": "https://patellaw.com/#firm",
"name": "Patel Injury Law",
"url": "https://patellaw.com",
"telephone": "+1-602-555-0143",
"image": "https://patellaw.com/images/office-exterior.jpg",
"priceRange": "$$",
"address": {
"@type": "PostalAddress",
"streetAddress": "1200 N Central Ave, Suite 800",
"addressLocality": "Phoenix",
"addressRegion": "AZ",
"postalCode": "85004",
"addressCountry": "US"
},
"geo": {
"@type": "GeoCoordinates",
"latitude": 33.4598,
"longitude": -112.0738
},
"areaServed": ["Phoenix", "Scottsdale", "Tempe", "Mesa", "Maricopa County"],
"openingHoursSpecification": [
{
"@type": "OpeningHoursSpecification",
"dayOfWeek": ["Monday", "Tuesday", "Wednesday", "Thursday", "Friday"],
"opens": "08:30",
"closes": "17:30"
}
],
"knowsAbout": ["Personal Injury", "Auto Accidents", "Wrongful Death", "Premises Liability"],
"aggregateRating": {
"@type": "AggregateRating",
"ratingValue": "4.9",
"reviewCount": 187
}
}
The Person block for an attorney bio page links back to that LegalService through the @id reference. That cross-link is how the model knows Sarah Patel works at Patel Injury Law and not at some other entity with a similar name.
{
"@context": "https://schema.org",
"@type": "Person",
"@id": "https://patellaw.com/attorneys/sarah-patel/#person",
"name": "Sarah Patel",
"jobTitle": "Founding Partner",
"image": "https://patellaw.com/images/sarah-patel.jpg",
"worksFor": { "@id": "https://patellaw.com/#firm" },
"alumniOf": [
{ "@type": "EducationalOrganization", "name": "Stanford Law School" },
{ "@type": "EducationalOrganization", "name": "Arizona State University" }
],
"hasCredential": [
{
"@type": "EducationalOccupationalCredential",
"credentialCategory": "Bar Admission",
"recognizedBy": { "@type": "Organization", "name": "State Bar of Arizona" },
"identifier": "294612"
}
],
"knowsAbout": ["Personal Injury Law", "Wrongful Death Litigation", "Insurance Bad Faith"]
}
The FAQPage block for a practice-area page wraps the question-and-answer section. Two important rules: only mark up content that is actually visible on the page, and keep answers between 40 and 250 words. Anything shorter looks like keyword stuffing, anything longer rarely gets pulled cleanly.
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [
{
"@type": "Question",
"name": "How long do I have to file a personal injury claim in Arizona?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Arizona's statute of limitations for personal injury is two years from the date of the injury under A.R.S. 12-542. Claims against a public entity have a 180-day notice requirement. Wrongful death claims also fall under the two-year window but start from the date of death."
}
}
]
}
The Google FAQ rich result rollback (and why FAQPage still matters)
Anyone who built schema for law firms before 2024 remembers when FAQPage markup produced expandable Q&A widgets directly in Google search results. Those rich results were the single most visible payoff for adding schema.
In August 2023 Google announced that FAQ rich results would only show for “well-known, authoritative government and health websites.” By early 2024 the change was effectively total. No law firm site qualifies. The FAQ rich result on Google is dead for legal.
That kill is the single biggest reason law firms now under-invest in FAQPage schema. The reasoning goes: if Google does not show the rich result, why bother. The answer is that the rich result was always a nice-to-have. The real value of FAQPage schema in 2026 is AI citation, not Google SERP enhancement. ChatGPT, Perplexity, and Google AI Mode read FAQPage markup as a signal that this content is structured Q&A and pull it directly into their answers. Sites that added FAQ blocks with schema saw a 44 percent jump in AI search citations across measurable categories in 2025-2026 studies. Featured snippets, which are still alive and widely served, also draw heavily from FAQPage markup.
The takeaway: keep shipping FAQPage schema even though Google killed the rich result. The audience for it is now the AI layer, not the blue links.
The order to ship schema on a law firm site
For a firm starting from zero, the rollout order matters. Each step compounds.
Step one is the LegalService block on the homepage. This is the entity anchor for the entire site. Get it right once and every other piece of schema can reference it through @id.
Step two is Person schema on every attorney bio page. Each Person links to the LegalService through worksFor. This is where AI engines learn which lawyer at your firm handles which case type.
Step three is FAQPage on every practice-area page. Pick the five to ten questions a real prospect actually asks during an intake call. Not the questions a content writer thinks they ask. Real ones.
Step four is AggregateRating and Review on the LegalService entity, pulled from your real review sources. If you have 187 Google reviews at 4.9 stars, that is what shows up. Do not invent or pad numbers.
Step five, only if you have multiple offices, is nested LegalService entries for each location, with their own addresses, phone numbers, and opening hours.
After step five the firm is into the long tail: BlogPosting schema on individual blog posts (helps AEO citation of educational content), VideoObject on attorney intro videos, BreadcrumbList for site structure. Useful, not foundational.
How to validate schema before you ship
Three tools, in this order:
Google’s Rich Results Test (search.google.com/test/rich-results) confirms whether a page is eligible for any rich result Google still supports. For law firms, that mostly means LocalBusiness enhancements. Run every page through it after you ship.
Schema Markup Validator (validator.schema.org) catches schema.org spec errors that the Rich Results Test does not flag. Use this for the deeper structural check.
Then a manual review in Google Search Console under Enhancements. Schema errors that survive both validators sometimes appear here a few days after Google re-crawls. Fix them before they multiply.
A common gotcha: Google’s tools will flag Attorney schema as deprecated but will still pass it. Do not interpret a passing validation as endorsement. Strip Attorney and replace with Person. Same for any property that schema.org has marked as Pending or Superseded.
Frequently asked questions about legal schema markup
Does schema markup directly improve Google rankings?
Schema does not directly affect rank position. It affects eligibility for rich results, knowledge panels, local pack inclusion, and AI citation. The downstream effect on traffic and lead volume is real, but it routes through visibility surfaces rather than the ranking algorithm itself.
Should we use Attorney schema or Person schema for our lawyer bios?
Person schema. Attorney was deprecated by Google in 2024 and produces no enhancements. Person schema with hasCredential pointing to bar admissions gives the same semantic meaning and is the current recommended pattern.
Can we add FAQPage schema to a page that does not have visible FAQs?
No. Google’s policy is that schema must mark up content that is visible to users. Marking up hidden FAQ content is a structured data violation and can trigger a manual penalty. If you want FAQPage schema, ship visible FAQ blocks.
How long until schema markup affects AI citations?
Most AI engines re-index content on a 7 to 30 day cycle. Once schema is live and validated, expect detectable citation changes within 30 to 60 days, with the full effect compounding over a quarter as the engine builds confidence in the entity record.
Do we need a developer to add schema, or can it go through a plugin?
Plugins are fine for static fields like firm name, address, and hours. For dynamic fields, especially Person schema on bio pages and FAQPage schema tied to specific content, hand-rolled JSON-LD or a custom build through your CMS produces cleaner output. WordPress plugins like Schema Pro or RankMath cover 70 percent of the work. The other 30 percent benefits from a developer.
What to do this week
If your firm has zero schema, ship LegalService on the homepage and Person on every attorney bio. That alone catches you up to roughly the 60th percentile of law firm sites in 2026. Then add FAQPage to your top three practice-area pages. The whole package, for a 5-attorney firm, is a one-day developer engagement.
If you want a free audit of your firm’s current schema implementation, plus a side-by-side comparison of how your site shows up in ChatGPT and Perplexity versus your three closest local competitors, get in touch with us or run our AI search ROI calculator to see what citation visibility is worth in your practice area.
Schema is the cheapest meaningful AEO win available to a law firm in 2026. Most firms still have not done it. That is the opportunity.
Tagged