SEO (search engine optimization) is the work of getting a law firm's website to rank in traditional search results. GEO (generative engine optimization) is the work of getting that same firm mentioned or recommended inside answers generated by tools like ChatGPT, Google Gemini, and Perplexity. They overlap far more than they compete.
What's the difference between GEO and SEO for law firms?
SEO and GEO answer two different questions. SEO asks: when someone types a search phrase, does this firm's page appear near the top of the results? GEO asks: when someone asks an AI tool a question, does that tool mention or recommend this firm in its answer?
The two are not rival strategies fighting for a firm's budget. GEO depends on most of the same raw material SEO does: a website that is easy to crawl, content that is accurate and current, and enough signals across the web that a firm is a credible source on its subject. A firm that has neglected its SEO fundamentals has almost nothing for an AI tool to draw on either.
What does traditional SEO optimize for?
SEO is the practice of making a website easy for search engines to find, read, and match to what a searcher typed. That includes technical basics (a site that loads reliably and that search engines can crawl), on-page content that matches the words and intent behind common searches, and off-site signals like other reputable sites linking to or mentioning the firm.
Success in SEO is measured by position: does the firm's page show up on the first page of results, and ideally near the top, for the searches its potential clients actually run. It is a ranking exercise. One page either outranks another or it doesn't.
What does GEO optimize for?
GEO is the practice of shaping a firm's content and online presence so that generative AI systems can find it, understand it, and choose to reference it when composing an answer to a user's question. Instead of competing for the top slot on a results page, a firm is competing to be one of the sources an AI tool decides is worth citing or naming.
This matters because of how these systems actually work. Google's own documentation describes AI Overviews and AI Mode as sometimes using a technique it calls query fan-out: the system runs multiple related searches behind the scenes, across different subtopics and sources, rather than matching a single exact phrase the way a classic search does. That means a firm isn't just competing for one query. It's competing to show up across a cluster of related questions an AI system generates on its own.
How does Google's AI Overview actually decide what to show?
According to Google's own Search Central documentation, AI Overviews and AI Mode are built on the same Search infrastructure used for classic results, not a separate system with its own rulebook. Google states plainly that there are no additional requirements or special optimizations needed to appear in these features, no new machine-readable files or AI-specific text files to create, and no special schema.org markup required beyond what standard Search already uses.
Crawling access is still controlled the normal way: through robots.txt directives aimed at Googlebot, since Google treats AI as integral to Search rather than a separate crawler to manage. And an AI Overview doesn't appear on every query. Google says it only shows one when its systems determine it adds something beyond the classic results, which means many searches, including plenty of legal ones, won't trigger one at all.
Does building GEO visibility mean giving up on SEO?
No, and treating GEO as a separate technical track is a mistake. Google's confirmation that no additional AI-specific setup exists, no special files, no extra schema, means the crawlable, accurate, well-organized site that supports good SEO is the same asset an AI system draws on when it looks for something to cite.
The practical implication is sequencing, not substitution. A firm doesn't choose between SEO and GEO. It builds one solid, accurate, technically sound website and set of content, and that work supports both a ranking on a results page and a mention inside an AI-generated answer.
| | SEO | GEO | |---|---|---| | What it optimizes for | Ranking position on a search results page | Being referenced inside an AI-generated answer | | How success is measured | Where a page lands for a given search phrase | Whether an AI tool mentions or names the firm | | What the underlying work is | Crawlable site, matching content, external links | The same crawlable, accurate content, read by an AI system instead of ranked by one | | Special technical setup required | Standard Search practices | None beyond standard Search practices, per Google's own documentation | | Who controls the outcome | Search engine's ranking algorithm | Each AI company's own system, on its own terms |
Do law firms need special schema markup for AI search?
Structured data is a standardized way of labeling the content on a page so a computer program, not just a human reader, can tell what it's looking at. Google's Search Central documentation describes it as a format for describing and classifying page content, and recommends a specific way of writing it, called JSON-LD, as the easiest to implement and maintain at scale. JSON-LD is a block of code placed in a page's markup that spells out facts about the page in a structured way.
Schema.org, the shared vocabulary that JSON-LD draws from, defines a LegalService type specifically for firms like this. It lets a firm specify details like opening hours, a price range, and a legal address that can differ from the address where the firm actually operates. To be eligible for any enhanced display in Google Search, all of the required properties for a given type have to be included.
None of this is an AI-only requirement. It's standard structured data that supports how Google displays results generally. A firm that has it in place has simply made its own facts easier for any system, human or machine, to read correctly.
What is llms.txt, and do law firms need one?
llms.txt is a proposed plain-text file some websites now publish at a specific address on their site. The idea behind it, as described in the specification itself, is that AI systems have a hard time processing a full, complex HTML page: their context windows (the amount of text a system can take in and consider at once) are often too small for an entire website, and converting HTML into clean, readable text is imprecise. llms.txt is meant to offer a simpler, markdown version of a site's key content that's easier for an AI system to read, whether at the moment it's answering a question or, potentially, during training.
The specification is explicit that the file only requires one thing: a top-level heading with the site or project's name. Everything else is optional. It's also explicit that this is a different tool from robots.txt, which controls what automated crawlers are allowed to access, rather than providing readable content.
This is a young, optional convention, not an established standard, and it isn't listed in Google's own documentation as something required or even recommended for appearing in its AI features. A firm can reasonably decide it's not a priority yet.
How should a law firm prioritize SEO and GEO work?
Start with the fundamentals, not the acronym. A website that loads reliably, that search engines can crawl without trouble, and that states the firm's practice areas, locations, and credentials accurately is the foundation both SEO and GEO stand on. Skipping straight to a tactic like llms.txt or extra schema without that foundation in place accomplishes very little.
Once the fundamentals are solid, the next step is finding out what AI tools currently say, rather than guessing. Since each AI company decides what its own tool surfaces, and that can change over time and differ across ChatGPT, Google Gemini, and Perplexity, the only way to know where a firm actually stands is to ask them directly. A free AI visibility check shows what these tools say about a firm today, which is a more useful starting point than assuming a strong Google ranking already covers it. From there, effort can go toward whatever gap the check actually reveals, rather than toward every tactic a marketing page recommends. For a deeper look at the practice of shaping content for these answers specifically, see our guide to generative engine optimization and how it relates to the narrower discipline of answer engine optimization.