The legal marketing industry has a pattern. A real technological shift happens, and within six months the conference circuit is full of agencies selling it as a completely new discipline that requires a completely new budget. Voice search got this treatment. Mobile-first indexing got it. Now answer engine optimization is getting it, and the pitch sounds the same as it always does: everything you did before is obsolete, hire us to start over.
Except the data doesn’t support the “start over” part. Studies covering thousands of ranking pages tell a boring story: the signals that predict traditional law firm SEO performance and the signals that drive AI visibility are mostly the same signals. Call it 80% shared infrastructure. The other 20% matters, but it’s not what’s filling up the webinar calendar.
The keyword era never ended. It just got a roommate.
There’s a popular version of this story where keywords died sometime in 2024 because people started talking to ChatGPT in full sentences instead of punching fragments into Google. That version is wrong.
Full-sentence queries have been showing up in Google search logs for years. Nobody invented a new way to ask questions. “What should I do after a car accident” has been a search query since before ChatGPT existed. The thing that actually changed is what happens on the other end: an AI can now grab that question, pull from a dozen sources at once, stitch together a paragraph-length response, and occasionally name a specific firm to call. Same input. Different output.
For law firms, the keyword research that drives SEO and the keyword research that drives AEO are the same research. Someone asking ChatGPT about personal injury law in Houston is asking the same thing they’ve been typing into Google since 2015. Law firms with practice area pages built around those queries, pages that give real answers instead of dancing around them, are already showing up in both channels without doing anything differently.
Where it diverges is on the delivery side. Google hands back a ranked list of links. ChatGPT hands back a finished answer. That changes how content needs to be structured on the page, but the underlying keyword strategy doesn’t need to be touched.
The trust shift is where it gets interesting
If there’s one thing that genuinely separates AEO from traditional SEO, it’s not a technical signal. It’s trust.
Traditional SEO lets you rank through brute force. Stack enough backlinks, nail the keyword targeting, keep the technical health clean, and a page can reach the top of Google without anyone confirming the business behind it is real. That’s how lead generation companies with no actual attorneys on staff end up ranking for “best personal injury lawyer.”
AI answer engines can’t afford to work that way. When ChatGPT recommends a law firm to someone asking about a custody dispute, it’s putting its own credibility on the line. If the recommendation turns out to be a lead-gen funnel or a defunct practice, the user loses trust in the platform. So these systems have built-in incentives to verify before they recommend.
What that looks like in practice: the AI cross-references your website against legal directories, bar association profiles, review platforms, news mentions, and professional listings. The question it’s trying to answer isn’t just “does this firm have a website.” It’s “does the information on this website match what Avvo says, what the state bar says, what Google reviews say, and what the local business journal wrote about them last year.”
That process has a name in the AEO world: entity verification. And the strange thing about it is that the data sources involved are all sources local SEO practitioners have been managing for a decade. Google Business Profile. Avvo. State bar records. Martindale-Hubbell. Yelp. Every local SEO checklist from 2016 included these platforms. The difference now is that AI platforms use them as a trust gate, not just a ranking signal.
Firms that kept all those listings accurate and consistent for local SEO purposes accidentally built exactly the trust infrastructure that AEO requires. Firms that ignored citations and focused only on backlinks are now finding out that ChatGPT won’t recommend a business when half its directory listings show a different phone number than the website.
The content architecture overlap is almost total
Topic clustering, internal linking, pillar page structures. These have been core SEO strategies for so long that they barely register as novel. But they’re doing more work than ever in an AI context.
When you analyze which law firm pages get cited by AI answer engines versus which ones get ignored, the pattern is consistent. A standalone page on car accident law, no matter how long, doesn’t carry the same weight as a cluster of 15 pages on the same topic, each covering a different angle, all linked together. Comparative fault gets its own page. Statute of limitations gets its own page. What to expect at a deposition, how settlements are calculated, when to file, what happens with uninsured drivers. That kind of depth tells a retrieval system the site actually knows the subject rather than just mentioning it once.
The Google antitrust trial produced court documents that help explain why this works. Google’s AI Overview system runs on something called FastSearch, powered by a signal set called RankEmbed. What RankEmbed does, stripped of the jargon, is measure how well the content on a site maps to the meaning behind a query. That’s a fancy way of saying it evaluates how thoroughly your site covers a topic. Content architecture is how you build that signal, and the approach is identical for SEO and AEO.
What the data says about the signals that don’t matter
Here’s where things get uncomfortable for agencies that have been selling PageSpeed audits and Domain Authority improvements as core deliverables.
Custom Legal Marketing published a study covering 1,750 search results across 50 U.S. metro markets. The Pearson correlation between PageSpeed scores and ranking position came back at r = -0.0705. That number is close enough to zero that you could round it there without anyone objecting. To make it concrete: 64.7% of the pages holding Position 1 in that dataset scored “Poor” on Google’s own Largest Contentful Paint test. The fastest page in the study ranked fifth. A page scoring 28 out of 100 ranked first.
A separate CLM study looked at Domain Authority across 12,794 results. Same story: r = -0.0819. Negligible.
If these metrics can’t predict where a page ranks in regular Google results, they’re certainly not predicting AI visibility. ChatGPT doesn’t load your page in a browser and time how long it takes. It reads text from an index. Your site could take nine seconds to render and the AI wouldn’t know or care. And Domain Authority is a score Moz invented. No AI platform uses it as an input.
Yet agencies continue to sell both as primary optimization levers. For SEO and AEO alike, the data says these are distractions from the signals that actually predict performance: content clarity, topical depth, entity consistency, and verifiable reputation.
Where the two actually diverge
The real differences between SEO and AEO are narrower than the marketing suggests, but they’re not nothing.
In SEO, ranking is a spectrum. Position 1 is better than Position 5, but Position 5 still gets traffic. In AEO, inclusion is binary. The AI either mentions your firm or it doesn’t. There is no Position 7 in a ChatGPT response. That raises the stakes on content structure because the AI is selecting one or two sources, not ten.
The other real difference is how the content gets used. Google looks at a page and decides where to put it in a list. An AI answer engine looks at a page and decides whether there’s a specific passage worth ripping out and dropping into a response. Pages that put the answer up front, clear and direct in the first couple of sentences, are the ones that get extracted. Pages that spend three paragraphs warming up before saying anything definitive get skipped, even if they rank number one in organic search.
And measurement is legitimately new territory. SEO has mature tracking: rankings, traffic, click-through rates. Tracking AEO means figuring out whether ChatGPT, Perplexity, Google AI Overviews, and Claude are actually mentioning a firm for specific queries in specific cities, and doing that across all of those platforms at the same time. The monitoring infrastructure for that kind of tracking is still being built. Most of the tools didn’t exist two years ago.
The implication nobody wants to hear
If SEO and AEO are 80% the same discipline, that means most law firms don’t need a new strategy. They need to do their current strategy better.
The firms that show up in AI-generated answers didn’t get there by hiring a boutique AEO shop and starting from scratch. They got there the slow way: years of building out real topical coverage, keeping their business information consistent everywhere it appears, writing content specific enough that a machine can verify it against outside sources, and earning the kind of reputation that doesn’t fall apart when an algorithm checks the receipts.
That’s not a revolutionary new playbook. That’s just good marketing, the kind that works regardless of which system is evaluating it.
