From the desk
Original writing on AI search, generative engine optimization, and law-firm marketing — by Shawn Lai. Each piece grounded in observed evidence, not opinion. New articles publish as the methodology produces them.
Browse by topic
Immigration is not a generic GEO engagement — it has multilingual search, community entity signals, and a sharper compliance line than any other practice area. Six questions to ask any agency before you hire, and the answers that separate an immigration-ready GEO firm from a generalist.
Billions now get answers from AI, and the click is vanishing into the answer. Why the shift to AI search is repricing law-firm marketing — and why being the cited answer now lifts both your organic exposure and your paid clicks. The 2026 data, with the caveats attached.
Wondering whether your firm shows up when clients ask an AI assistant for a lawyer? Ten questions you can answer yourself — each mapped to something AI engines actually check — that tell you where your firm's AI-search foundation is solid and where it's quietly costing you referrals.
Every legal marketing agency now says it does AI search. Most are running an SEO playbook with a new label. Here's what genuinely differs between the two disciplines — the corpus, the unit of success, the measurement, and the compliance surface — and the questions that tell you which one you're actually buying.
The GEO category is new and every provider claims to do "AI." Seven concrete tests that separate a real GEO partner from a repackaged SEO retainer — measurement you can check, built-in compliance, legal-specific expertise, honest mechanics, low-risk terms, visible work, and self-honest claims.
In AI search the gap that matters isn't where you rank — it's whether you're cited at all. Pages cited inside a Google AI Overview earn roughly 2.1% click-through vs about 0.9% for uncited pages on the same results page (Seer Interactive, 2026). AI Overviews now appear on more than one in five US searches, and more on the questions that lead people to a lawyer.
SEO checklists tell law firms to add ten kinds of schema. For AI search, a small subset does the real work — and structured data is still lawyer advertising. What to prioritize, what to skip, and the compliance tripwires.
Most marketing optimizes what AI engines display. The smarter work optimizes what they retrieve. Entity consolidation is the single highest-correlation signal for AI citation rate — and the work most agencies don't do because it's invisible to traditional SEO dashboards.
The retrieval architecture in plain language, the five signals that actually move citation rates for legal queries, and what everyone fixates on that doesn't.
Most AI-visibility reporting is a screenshot, a traffic chart, or a percentage with no denominator. This is our full method instead — the six engines we track as separate rows, why the prompt set is frozen, why a person still runs a manual pass every month, the numbers we refuse to report, and how to run a lighter version of it yourself.
A week-by-week walkthrough of month one — the 50–100-prompt baseline, entity consolidation, the content-gap map, the compliance log, and the first PROOF Report. Specific enough to picture the engagement; specific enough to audit the method.
Investor and employment-based immigration is the high-value end of a community practice — and in AI search it's often won or lost on a single purpose-built page. Why the EB-5, E-2, and employment answer is winnable, what these clients ask the engines (often in their own language), and the extra compliance layer investment marketing carries.
Marriage green cards, adjustment of status, family petitions, and naturalization are the highest-volume, most-researched immigration work — and where clients are most likely to ask an AI first, often in their own language. Why family-based and citizenship work is the sub-lane to win first, what these clients ask the engines, and how a community firm becomes the named answer.
If your immigration practice serves Korean-, Vietnamese-, Chinese-, or Spanish-speaking clients, your language community is your strongest edge in AI search — a defensible, under-contested niche. How the engines decide who to name, why in-language content matters, and how H&H Law became the #1-named Korean immigration firm in its market.
Family law is high-volume, high-emotion, and searched privately — exactly the behaviour AI assistants were built for. The prompts that retain clients aren't the ones most firms optimize for, and the outcomes clients most want promised are the ones a bar rule forbids.
Estate planning is the one vertical where the buyer and the AI-search shift line up almost perfectly: the 45-and-older, higher-income clients planning wills and trusts are the fastest-growing adopters of AI assistants for important decisions. The prompts that convert are deliberate and jurisdiction-specific — and the marketing language firms lean on hardest is what a bar rule reserves.
Personal injury is the most contested vertical in legal marketing and the one where AI search behaves least like Google. The prompts that bring in cases aren't the ones most PI firms optimize for — and the words that read like good marketing are often the ones a bar rule prohibits.
A large share of the highest-intent immigration searches in the US are made in Spanish — and AI assistants answer in the language they're asked in. Most firms are invisible to that demand. Why machine translation doesn't fix it, the notario compliance trap, and what genuinely closes the gap.
Generic law-firm GEO advice misses what's specific about immigration: clients search in many languages, intent is urgent and high-stakes, and the law changes constantly. What actually gets an immigration firm cited by AI search — and the compliance tripwires that are sharper in this vertical.
In August 2026 a UK lawyer was struck off over AI-fabricated citations; the same month a top firm made a critical-AI-judgment seat compulsory for trainees. The lesson for hiring a GEO or AI-marketing agency: your website is lawyer advertising, and the vendor’s real value is the human accountability layer, not the generation. The one question to ask before you sign.
California isn't the strict one. If your firm advertises in more than one state, AI-generated content has to clear the strictest rule that applies — filing regimes, mandated disclaimer wording, and reserved words that AI reaches for by default.
The most-referenced and least-quoted document in legal-vertical AI marketing. The guidance is more specific than the industry's general interpretation suggests — and the compliance work that protects the firm also lifts its AI-search citation rate.
How AI-generated content can quietly cross state-bar advertising-rule lines, why generalist GEO agencies don't catch the violations, and the practitioner checklist for catching them before they ship.
Avvo, Justia, FindLaw, Martindale, Google Business Profile, the state bar — ranked by what actually moves AI-search citations, not backlinks. Plus the consistency rule that beats any single listing, and the ratings-and-badges compliance trap.
Avvo, Justia, JD Supra, Google Business Profile, and state/local bar. Five citation sources AI engines disproportionately pull from for legal queries — ranked by leverage, with operational checklists.
A quantitative study of 200 California firms across 4 AI engines, 60 days, 50 buyer-intent prompts. Pre-registered methodology, publishable either way. Q4 2026.
We pre-registered a timeline for PROOF Series #1 and committed to disclosing any deviation over 30 days. The deviation happened. Here is the disclosure: data collection moves to September 1 – October 30, publication to Q4 2026. Nothing about the methodology changes.
Ready when you are
Free AI Visibility Audit. 60 seconds. See your citation gap before your competitors do.