One year into AI Overviews being the default for most legal queries, here’s what actually happened to law firm SEO. The informational-query traffic vanished, mostly for good. The buyer-intent traffic survived — in some practice areas it got more valuable, not less. The firms getting cited inside AI Overviews mostly weren’t optimizing for them. The split between consideration content and intent content is now the whole game.
I’ve been watching this across a portfolio of client firms — personal injury, criminal defense, family, estate, with a smaller mix of business law. The data isn’t the same in every practice area, and the firms that responded fast look very different at month twelve than the firms that ignored the shift. This is the honest read.
What AI Overviews actually do for legal queries now
For most informational queries, the AI Overview now sits at the top and answers directly. “What is comparative negligence in Arizona,” “how long do I have to file a personal injury claim in Phoenix,” “what’s the difference between Chapter 7 and Chapter 13” — the user reads three to six sentences, sees citations collapsed behind an expansion, and most never click. The blue link is scrolling territory.
For buyer-intent queries — “best personal injury lawyer Phoenix,” “DUI attorney near me,” “Scottsdale estate planning attorney” — the AI Overview usually doesn’t appear, or it’s a stripped-down list of firms with one-line summaries. The local pack is still there. Organic results still get clicks. Behavior on these is essentially what it was eighteen months ago.
For the queries in between — “do I have a case for X,” “should I hire an attorney for Y,” “what does a divorce cost in Maricopa County” — the picture is mixed. Sometimes the AI Overview answers and the user converts inside the AI Overview to a recommended firm. Sometimes it teases an answer the user clicks through to verify. Practice area matters here. Personal injury behaves differently from estate planning, which behaves differently from criminal defense.
The informational query massacre
If your firm’s blog was built on the “answer every legal question a layperson might ask” model — and most are — your traffic is down somewhere between forty and eighty percent compared to where it was before AI Overviews defaulted on. The worst drops we audited this year were at firms whose vendors had been pumping out forty blog posts a month targeting informational long-tail queries. Every one of those posts now competes with an AI Overview that answers the question better, faster, and without making the user click.
The blog traffic that survived has a pattern. Questions where the user needs more than three sentences — edge cases, procedural detail, jurisdiction-specific answers the AI summarizes badly. Local terrain — Maricopa County Family Court procedures, Phoenix-specific PI thresholds, Arizona estate planning forms. The geographic and procedural specificity makes the AI Overview less useful, so the user still clicks through. For why volume blog content stopped working, see will blogging help my law firm rank.
The hardest hit is the generic “what is” post with a Wikipedia-grade answer. “What is negligence.” “What is probate.” “What does a personal injury attorney do.” These pages are not coming back. The firms that built their content programs on them are paying retainers to vendors whose monthly outputs are functionally subtracting from revenue. The retainers haven’t been renegotiated. They should be.
Which firms get cited in AI Overviews — and why
Across our portfolio and the audits I’ve run on competitors, the firms cited inside legal AI Overviews are usually not the firms with the most blog content. They’re the firms with the cleanest, most structured answer pages — often inside their practice area pages. The cited page tends to have a clear answer in the first paragraph, headings that map to the question precisely, and schema markup that signals what the page is for.
Three other patterns. Cited firms tend to have substantial review counts and strong GBPs in the relevant locality — the AI weights real-world legitimacy signals heavily. Cited firms usually have an attorney byline or author markup, not anonymous “Posted by Admin.” Cited pages live on the firm’s main domain, not a microsite or content farm subdomain.
The fundamentals are unchanged from pre-AI ranking — clarity, relevance, trust, structure. The schema layer matters more than it used to because the AI is parsing it aggressively. Schema markup for law firms has been the highest-ROI technical fix on our portfolio this year, by a comfortable margin. Not because schema is magic — because the AI Overview is parsing it.
The firms getting cited in AI Overviews aren’t the ones who optimized for AI Overviews. They’re the ones who already had clean, specific, structured answer pages — and then the AI started reading them.
The consideration vs intent split
The biggest mental-model shift the year forced is the split between consideration content and intent content. They behave like different products now. AI Overviews made the gap impossible to ignore.
Consideration content is the informational layer. “What is X,” “how does Y work.” The user is trying to understand. They might never hire a lawyer, or might in eighteen months. The content earns traffic but rarely a phone call. After AI Overviews, most consideration traffic is gone — and what’s left is worth less, because the AI already answered the basic question before the user clicked.
Intent content is the conversion layer. The practice area page for “Phoenix DUI attorney.” The location page for “personal injury lawyer Scottsdale.” The lead attorney’s bio. The reviews page. These are the pages a user reaches when they’ve decided to hire and are evaluating candidates. AI Overviews barely touch this terrain. Traffic intact, conversion rates steady, value per visit slightly up because visitors arrive further along in the decision.
The implication is uncomfortable for anyone selling blog volume. Spending budget on the intent layer — better practice pages, better location pages, better bios, better trust signals — is now obviously more valuable than the consideration layer. The blog isn’t dead, but it has to do something different. See practice pages vs blog content for the longer treatment.
What this means for content strategy going forward
If I had to give a firm one piece of guidance for the next twelve months — stop publishing informational blog content at volume, and put the recovered budget into making your existing practice area pages the best version of themselves on the internet. Most firms are sitting on eight to twenty practice pages that haven’t been seriously rewritten since 2022. Those pages, fixed, pay back ten times what another forty AI-generated blog posts will.
Second — build the answer-page layer the AI Overview wants to cite. A small number of deeply specific, well-structured answer pages on the questions prospective clients ask before hiring. Not “What is personal injury.” More like “How long does a Maricopa County personal injury case take.” Clean schema, attorney attribution, direct answer in the first paragraph, structure the AI can parse. Done well, the firm becomes one of the cited sources inside the AI Overview that displaced its old blog traffic.
Third — review velocity and GBP health are now table stakes for AI citation visibility, not just local pack ranking. The AI reads social proof at a depth the old algorithm didn’t. Firms with thin review profiles aren’t getting cited even when on-page content is strong.
The question I get from owners more than any other this year — is SEO still worth it. The honest answer is yes, but the version that’s worth it is different than the version most have been buying. Fixing the intent layer, building a small high-quality answer layer, keeping reviews and GBP healthy — that’s the work. Pumping out blog content at volume, chasing informational queries, paying for vanity backlinks — that work is largely over. The fuller argument is at is SEO still worth it with AI search. One year in, the firms that responded fast cut their blog programs, redirected budget into practice page rewrites and schema, and kept organic case volume stable or growing. The non-responders are watching traffic decline quarter over quarter and renewing the same retainer producing the decline.
— The owner, PHX Search Co.
