Service
AI Search Optimization
Structured data, entity signals and answer-first content so AI can verify and cite you.
Entity consistency
ExampleThe three most common first-audit findings: a directory that added words to the name, one carrying an old phone number, and an unclaimed listing.
Last updated: September 2026
What is AI search optimization for a law firm?
AI search optimization is the set of changes that make a law firm easy for AI engines to verify and quote: structured data that states who you are and what you do, identical name and contact details across directories, pages that answer client questions directly, and permission for AI crawlers to read your site.
AI assistants do not rank pages. They decide whether an entity is a credible answer. AI search optimization, often called generative engine optimization or GEO, is the work of making your firm easy to verify: who you are, where you practice, what you handle, and what others say about you, expressed consistently everywhere a model looks.
What is included
- LegalService, ProfessionalService and LocalBusiness schema on every relevant page, with sameAs links to your profiles
- FAQPage and Article schema on content, Person schema for each attorney
- Entity alignment across your site, Google Business Profile, the State Bar, Avvo, Martindale, Justia and Yelp
- robots.txt and llms.txt configured for GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot and Google-Extended
- Answer-first rewrites of your core practice pages: question headings, 40-60 word lead answers, then depth
- Monthly citation tracking across four assistants for your query set
How the work runs
- 1
Entity map
Every place your firm is named, with every inconsistency flagged.
- 2
Schema
Typed structured data on every page, validated, with the deprecated Attorney type replaced.
- 3
Answer-first pages
Your practice pages rewritten so the first paragraph is the answer a model would quote.
- 4
Crawler access
AI bots allowed, sitemaps clean, no rendering traps.
- 5
Track and iterate
Citations logged monthly; the query set grows as you win.
What is entity alignment and why does it come first?
Entity alignment is making every place the firm is named say the same thing: legal name, address, phone, website, practice list, languages and attorney names, identical on the State Bar, Google, Apple, Bing, Avvo, Justia, FindLaw, Martindale, Yelp and LinkedIn. AI engines cross-check these before naming a firm, and a mismatch reads as two firms with half the signal each.
Entity consistency
ExampleThe three most common first-audit findings: a directory that added words to the name, one carrying an old phone number, and an unclaimed listing.
| Field | Rule | What we fix most often in LA |
|---|---|---|
| Firm name | One legal name, same suffix everywhere | 'APC' on some listings, 'A Professional Corporation' on others |
| Address | One format, suite on the same line, no old office | Glendale office on Google, Downtown on Avvo |
| Phone | One number, no call-tracking numbers on directories | Tracking numbers that split the signal |
| Practice list | Same list, same order | 'Personal injury' left over from 2019 |
| Languages | Same on Google, Avvo and the site | Spanish on the site, nothing on the profiles |
The twenty listings and the order to fix them are in the directory consistency checklist. Alignment is a working day for a solo firm and a week for a multi-office firm, and it is the prerequisite for everything after it.
What schema do we deploy?
LegalService and ProfessionalService for the firm, with one identifier reused on every page and sameAs links to every aligned profile; Person for each attorney linked to the firm; FAQPage on every question block; Article on guides with a named author; BreadcrumbList on inner pages. Validated after every deploy, because a broken block is silently ignored.
| Type | Pages | What it tells an engine |
|---|---|---|
| LegalService + ProfessionalService | All | Who the firm is, where it practices, which profiles are the same entity |
| Person | Attorney profiles, bylines | Who is responsible for the content and where else they exist (State Bar, LinkedIn, Avvo) |
| FAQPage | Question content | Question and answer pairs to extract, matching the visible text |
| Article | Guides, articles | Author, publisher, dates; freshness is a citation signal |
| BreadcrumbList | Inner pages | Where the page sits in the site |
We do not use the deprecated Attorney type and we never mark up self-written reviews. The copy-ready JSON is in the schema guide, and the schema generator produces the firm and attorney blocks in the browser if you want to see yours first.
What does an answer-first rewrite change on a practice page?
The first paragraph. A question-phrased heading and a 40-to-60-word answer replace the firm history that usually opens the page, followed by process, cost, timeline, courthouse and FAQ. About 55% of AI citations come from the first 30% of a page, so where the answer sits decides whether the page is quoted.
| Before | After |
|---|---|
| 'Founded in 2009, our firm has served Glendale families with compassion…' | 'How is child custody decided in California? California courts decide custody on the child's best interest…' (45 words) |
| Practice area described in general terms | Process steps, LA County timeline, fee range, courthouse the matter is heard in |
| 'Contact us' at the bottom | FAQ block of six to ten client questions with FAQPage schema, then the call to action |
| Unsigned | Named attorney with Person schema and a visible 'last updated' date |
Foundation includes the rewrite of the main practice page; Growth rewrites every practice and city page and adds the content cadence.
Which AI crawlers do we allow, and does it cost anything?
All the retrieval bots: OAI-SearchBot and ChatGPT-User, PerplexityBot and Perplexity-User, Claude-SearchBot and Claude-User, Google-Extended alongside Googlebot, Bingbot and Applebot. Blocking them removes the firm from AI answers. Firms that want to opt out of model training only block GPTBot, ClaudeBot and Applebot-Extended and keep the rest open.
- We read robots.txt, the CDN's bot settings and the security plugin, because each can block retrieval bots independently.
- We confirm practice pages return server-rendered HTML to those bots; most do not execute JavaScript.
- We ship llms.txt because it is ten minutes of work, and we say plainly that no engine has confirmed reading it.
The full user-agent list and a sample robots.txt are in which AI crawlers a law firm should allow.
How is progress tracked?
The query set from your audit becomes the tracked set. Every month it runs through the four engines and we report citation rate overall, by engine and by query group, alongside entity score, review velocity and the content shipped. Schema and entity fixes usually show within two to six weeks; content-driven citations take three to five months.
Citation rate
48 tracked queries
Engines citing you
ChatGPT · Perplexity · AIO
Entity score
from 31 at audit
Reviews this month
Google · Avvo
Citation rate · 12 wk
| Month | What moves first | What we watch for |
|---|---|---|
| 1 | Entity score to 100%; schema validated; crawlers confirmed | Firm named for branded queries in every engine |
| 2 | First unbranded citations on procedural and courthouse queries | Which engine moves first (usually Perplexity) |
| 3–5 | Core practice + city queries | Citation rate climbing; calls that mention an AI answer |
| 6+ | Compounding with content and reviews | 30–50% on the tracked set |
Outcome
Your firm becomes an entity models can confirm, which is the prerequisite for being named.
FAQ
AI Search Optimization: questions.
No. Classic technical and local SEO still runs underneath. AI search optimization is the layer on top that most agencies have not built.
It is cheap and harmless, so we ship it, but the evidence that any major AI engine reads it is thin. Schema, entity consistency and answer-first content carry the weight.
LegalService or ProfessionalService for the firm, Person for each attorney with sameAs links, FAQPage for question content and Article for guides. The Attorney type was deprecated and should not be used alone.
Schema and entity fixes are picked up within two to six weeks. Content-driven citations for competitive queries usually take three to five months.
Not if you want to be cited. Retrieval bots such as OAI-SearchBot and PerplexityBot are how you get named. We allow them and monitor what they fetch.
Still have questions? Talk to us →
Free citation audit
Find out who AI recommends instead of you.
Written report plus a thirty-minute walkthrough. No pitch until you ask for one.