What is AEO for law firms?
Answer engine optimisation (AEO) for law firms is the deliberate work of improving how accurately and prominently a firm is represented in AI-mediated discovery — when a prospective client asks an AI system which lawyer or firm to consider. It spans the firm’s website, lawyers, locations, content, technical accessibility, reputation and the third-party sources that describe it.
The exact relationship between those inputs and AI recommendations is still developing. That is why I treat AEO as a discipline that begins with measurement rather than with a list of tactics.
AEO, GEO and AI SEO: is there a difference?
Generative engine optimisation (GEO), AI SEO and LLM optimisation are overlapping labels for the same broad problem: how content and organisations are represented in answers produced by generative AI systems. For a law firm, the practical questions are the same whichever term is used.
I use these terms when they are useful, but I’m less interested in winning the naming argument than in understanding the behaviour underneath them. The terminology will probably keep changing; the question — does AI understand who your firm is, and does it consider you? — will not.
How is AEO different from SEO?
SEO is largely about being found and ranked in a list of search results. AEO is about how a firm is represented inside a synthesised answer — whether it appears at all, how it is described, and whether it is placed into the user’s consideration set or recommended.
SEO remains important: AI systems can use web search and web content, and a well-structured, crawlable, authoritative website helps both. But AI recommendation introduces a different output. A system can combine many sources and explicitly narrow the choice — which is why “ranking” is often the wrong mental model. I explain why in Why “ranking in ChatGPT” may be the wrong mental model.
How do AI systems recommend law firms?
From the outside, we can observe the stages: a client describes a problem, the system interprets intent, may retrieve information from the web, draws on sources about firms, resolves which firms are being discussed, and produces an answer that may list, compare or recommend. We cannot see every internal weighting.
That distinction — what is observable versus what is inferred — matters. It is the basis of the FirmRanker research programme and is explained in more depth in AI & Law.
What is the difference between an AI mention and a recommendation?
An AI mention and an AI recommendation are not the same thing. A firm can appear in an answer without being placed into the user’s consideration set.
I use a working framework of five categories — mentioned, suggested, recommended, source-only and negative/cautionary — and classify how the answer treats the firm, not how the question was worded. The full framework is in What actually counts as an AI recommendation?
How should law firms measure AI visibility?
A single ChatGPT response is an observation, not a ranking. Measuring AI visibility requires repeated observations across the prompts, models and markets that matter to the firm — and classifying whether the firm was mentioned, suggested or recommended each time.
A serious baseline covers commercially relevant practice areas, important markets, realistic prompts, multiple AI platforms, repeated observations, competitor visibility and source behaviour.
In practice
A single query is a demonstration. A structured set of repeated queries is measurement.
Start with the AI Visibility Checklist, then see how FirmRanker approaches measurement in the methodology.
Eight foundations I would examine
Entity clarity
Is the firm consistently named and described?
Expertise architecture
Can a person — or machine — understand what the firm is genuinely expert in?
People
Are lawyers clearly associated with their work, jurisdiction and evidence of expertise?
Evidence
Does the website substantiate its claims?
Third-party authority
What credible external sources describe or recognise the firm?
Technical accessibility
Can important information be efficiently retrieved and understood?
Reputation environment
What does the broader web say about the firm?
Measurement
Can the firm observe whether changes correspond with changes in AI visibility?
What I would not promise
If you are being sold any of these, ask for the evidence.
- Permanent ChatGPT rankings
- Guaranteed recommendations
- That schema alone solves AEO
- That publishing hundreds of AI-generated pages solves AEO
- That one directory placement causes recommendations
- That one successful prompt proves optimisation worked
I go through these claims in detail in AEO for law firms: evidence versus hype.
A sensible operating model
Measure → Diagnose → Improve → Re-measure → Learn. Measurement comes first because, without a baseline, a firm cannot tell whether anything it changes has made a difference.
My view
Measurement should come before optimisation. The firms that learn fastest may have an advantage over firms that simply follow the latest checklist.
— Dan Toombs