Answer Engine Optimization is the work of structuring a site so AI engines (ChatGPT, Perplexity, Claude, Gemini, Google AI Overviews, Microsoft Copilot) can quote it cleanly when a prospective client asks them a question. For most industries this is mostly a schema and content-depth problem. For law firms it is also a compliance problem, because the things that make a page citable (specific claims, concrete numbers, FAQ-style direct answers) are also the things that bar advertising rules constrain most carefully. The trick is doing the AEO work inside the envelope of Rule 4.06 in Ontario or its provincial equivalents.
The four AEO pieces that matter for law firms
1. LegalService schema, one per practice area
The base schema for a law firm is LegalService rather than the generic LocalBusiness most theme builders default to. LegalService carries fields for areaServed, serviceType, and provider, which the AI engines parse to understand what kind of legal work the firm does and where it does it. The discipline is one LegalService block per practice-area page, not one block for the whole site. A firm with personal injury, family, immigration, and criminal pages has four LegalService schemas, not one.
2. FAQPage schema with the questions clients call about
The questions that get cited in AI Overviews are usually FAQ-shaped. "How long does a divorce take in Ontario." "What happens at a settlement meeting in a personal injury case." "Can a permanent resident sponsor a spouse who is currently in Canada on a visitor visa." We pull the questions the firm's intake staff hear on the phone, write the answers within the firm's province's marketing rules, and mark them up as FAQPage. The answers cannot promise outcomes, cannot guarantee timelines, and cannot compare results to other firms in ways the bar would flag. They can be specific, plain-spoken, and useful, which is what AI engines weight.
3. llms.txt at the root
llms.txt is a text file at the root of the site that tells AI crawlers which pages matter, which pages to skip, and how to summarize the firm. For a law firm this is the place to lay out the practice areas, the offices, the lawyer names with bar admissions, the languages served, and the firm's basic positioning. We write it once during the build and keep it current as the site changes.
4. Per-page topic depth
AI engines weight depth over breadth. A single practice-area page that covers personal injury comprehensively (what counts as personal injury, how contingency fees work in Ontario, what the settlement timeline usually looks like, what happens at mediation, what evidence the firm needs from the client) will outrank ten thin pages with a paragraph of copy each. The Engine tier exists precisely for firms that need this depth crossed with city specificity (Toronto personal injury, Mississauga personal injury, Brampton personal injury) at scale.
The compliance gate
Every line of copy passes through a law-society-advertising-rule check before publication. For Ontario firms that means Rule 4.06 of the Rules of Professional Conduct, which prohibits marketing that is false or misleading, that compares the firm's services to other firms in ways that cannot be factually verified, that implies a guarantee, or that promises results. For BC firms that means the equivalent in the BC Code. For Quebec firms that means the Barreau du Quebec's rules on professional advertising. We read the relevant rule before drafting any headline.
The practical result is that the copy is plain-spoken, hedged where it should be hedged, and quantified only where the bar permits quantification. "We have handled personal injury matters in Ontario for over a decade" is permissible; "We are the top personal injury firm in Toronto" usually is not. The compliance discipline is not a constraint on AEO; it is a constraint on which AEO patterns are usable, and the usable ones are still enough to rank.
Specific examples
For a Toronto family law firm, the AEO work usually shapes up as: a family law landing page with LegalService schema, FAQ blocks on divorce timelines and custody under the Divorce Act, a lawyer profile page for the lead family lawyer with Person + JobTitle schema, an article layer on common questions ("how does property division work in Ontario") that links back to the practice page, and a Cite-met dashboard that shows which AI crawlers are reading the firm's pages and how that AI-search traffic grows month-to-month. For an immigration firm in Markham serving a Mandarin-speaking client base, the same model plus a parallel Mandarin page layer with hreflang markup.
The citation pattern
Once the schema, FAQPage, llms.txt, and per-page depth are in place, the AI-search traffic usually starts to show up in the Cite-met dashboard inside six to twelve weeks. GPTBot and PerplexityBot tend to pick up first; Google's AI crawlers tend to take longer because they pull from a wider index. We track AI-search traffic across six engines in real time, not quarterly. The studio's own site gets 32% of its traffic from AI engines and somewhere north of 11,000 visits in the last seven days came from those engines directly. The same pattern is reachable for law firms, with the compliance gate intact.
For the underlying mechanics of how AI engines cite pages, the longer piece is at what is answer engine optimization. For the schema work specifically, the technical breakdown is at the structured data guide. For the cite-met AI-search traffic tracking tool, cite-met.com is the studio's own dashboard. Other regulated-industry buyers like accountants and dentists work inside similar compliance constraints with the same AEO model.
Run the free audit and the studio will tell the firm where its current site sits on the four pieces above, what would need to change to get cited, and what the rebuild would cost. The audit is a one-page report, not a sales pitch.