Why your law firm doesn't appear in AI search - and what to do about it
When a prospective client asks an AI assistant "which Cape Town attorneys handle commercial lease disputes," the answer draws from a pool of structured, authoritative web content - not from a directory listing or a paid ad. Most South African law firms are not in that pool. This is a solvable problem, and it matters more with each month that AI search answers displace traditional blue links.
How AI search answers are generated
Large language models used in search (Google's Gemini for AI Overviews, Perplexity, ChatGPT, Claude) do not simply retrieve pages - they synthesise answers from content they have indexed and crawled. The content selected tends to share four characteristics: it directly answers a specific question, it signals authority through structure and citations, it is attributed to an identifiable entity, and it uses schema markup that search engines can parse.[1]
This is Generative Engine Optimisation (GEO) and Answer Engine Optimisation (AEO). GEO is about structuring content so it can be synthesised by AI. AEO is about directly answering the questions prospects ask.
Why law firm websites fail at AI search
No entity definition
AI search engines need to understand what your firm is. Schema markup using the LegalServiceand Organization types from Schema.org signals your firm's identity, location, practice areas, and contact details in a machine-readable format. Most law firm websites have none.
No question-answering content
AI systems favour content that directly answers the questions people ask. "What do employment lawyers charge in Johannesburg?" "How long does a property transfer take?" "What is the difference between a notarial bond and a mortgage bond?" If your website does not answer these questions, you cannot appear in the answers. FAQ schema (the FAQPage type) makes those answers machine-readable.
No author attribution
Google's E-E-A-T framework - Experience, Expertise, Authoritativeness, Trustworthiness - treats author attribution as a positive signal, especially in legal content. An article signed by a named, admitted attorney with a verifiable profile outranks the same article published anonymously.
No external citations or links
Citations from authoritative sources (legislation, LPC rules, court judgments) and inbound links from reputable legal directories signal that your content is part of the broader legal information ecosystem. This is exactly the kind of signal AI retrieval systems weight positively.
What to implement
| Element | What it does | Priority |
|---|---|---|
| LegalService schema | Identifies firm, location, practice areas to AI crawlers | High |
| FAQPage schema | Makes Q&A content machine-readable for featured answers | High |
| Article schema with author | Attributes content to named professionals | Medium |
| Q&A content pages | Directly answers the questions prospects search | High |
| Citations to primary sources | Signals authority and verifiability | Medium |
| Google Business optimisation | Entity confidence for local AI answers | High |
Frequently asked questions
Does this replace traditional SEO?
No - it complements it. Traditional SEO (keywords, backlinks, technical health) remains important for blue-link search. GEO/AEO extends that foundation to AI-generated answers. The two share most of the same inputs: quality content, clear structure, authoritative signals.
How long before results show?
Schema changes and content additions are typically indexed within days to weeks. Appearing in AI-generated answers depends on content quality, query match, and how often the AI's knowledge is refreshed - there is no fixed timeline, but firms with better-structured content consistently outperform those without it.