Category: AI Search

Ranking first on Google no longer means what it used to. Legal is one of the search categories most aggressively summarized by Google’s AI Overviews, and multiple independent analyses now show click-through rates to the top organic result dropping meaningfully whenever an AI Overview appears above it. The traffic didn’t disappear — it moved, and where it moved matters more than the raw percentage drop.

Quick answer

Being named inside an AI-generated answer is increasingly the first gate a prospective client passes a firm through, with a Google search to verify still happening as a second step for most people before they call..

Independent studies through 2026 put the click-through rate (CTR) reduction for a page-one result, when an AI Overview appears above it, somewhere between roughly 34% and 58%, with the deeper end of that range reported more recently — meaning the effect appears to be intensifying, not stabilizing. Legal search triggers AI Overviews at some of the highest rates of any category.

What the data actually shows

Reported figures vary by study and methodology, which is worth being upfront about rather than picking the most dramatic number:

• One widely cited analysis found the CTR reduction for the top organic result went from roughly -34.5% in early-to-mid 2025 to roughly -58% by early 2026 — the effect nearly doubling over about eight months.
• Estimates of how often AI Overviews (generative engine optimization) appear on legal queries range from roughly a quarter to more than three-quarters of legal searches, depending on how “legal query” is defined and which query types (informational vs. transactional) are included in the sample.
• Ranking in the organic top 10 doesn’t guarantee an AI citation — one analysis found only about 17% of AI Overview citations also appeared in the organic top 10, and a separate study found the overlap between the two fell substantially over the same period.

The range itself is the honest takeaway: the exact percentage depends on which study and which query set, but every credible source points the same direction — AI-generated answers are absorbing a growing share of what used to be a click.

What “the click moved” actually means

The behavior pattern showing up across recent legal-consumer research is a two-step search, not a vanished one: a prospective client asks an AI system for a recommendation or an answer, then verifies what they found with a traditional Google search before reaching out. Being mentioned inside the AI answer increasingly decides who makes the shortlist; the Google search and the firm’s actual site then decide who gets the call.

That means two different things now have to work:

1. Being cited by the AI system in the first place — which depends more on structure, sourcing, and third-party authority than on organic ranking position alone. Recent analysis has found that earned mentions on sites like Reddit, YouTube, and Wikipedia often outweigh a firm’s own blog as a citation source inside AI answers — an uncomfortable fact for firms that have invested heavily in owned content and little in earned mentions elsewhere.
2. Holding up under the verification click — reviews, case results, clear practice area pages, a site that loads fast and reads as credible once someone actually lands on it.

What doesn’t change

Organic search still drives the large majority of legal client acquisition clicks industry-wide, and AI Overviews appear less often on the highest-intent, transactional-style legal queries than on broad informational ones. The compression is real, but it’s concentrated at the top of the funnel — the “what do I do after X happened” queries — more than at the point someone is actually ready to hire.

FAQ

Should firms stop investing in their own blog content because AI is citing other sources instead?
No — owned content is still what AI systems and human readers verify against once a firm is in consideration. The gap is that owned content alone is no longer sufficient; earned mentions on third-party sites increasingly matter for the citation step itself.

Does a lower click-through rate mean less overall demand for legal services?
Not necessarily. Reported data suggests overall search demand hasn’t dropped — the same number of people are searching, but more of them get an answer without clicking through on the first query.

How would a firm know if it’s actually being cited by AI systems for its practice areas?
This requires checking directly across AI platforms since it isn’t visible in standard analytics — a growing category of tools exists specifically to monitor AI citation, separate from traditional rank tracking.

Publishing is the easy part. Most law firm content dies quietly the day it goes live, because nothing happens to it afterward — no one checks if it’s ranking, no one improves it, no one connects it to the pages already getting traffic. Here’s what the sequence should look like instead.

Quick answer

A blog post that’s actually working goes through four stages after it publishes: it gets written to a structural standard search and AI answer engines can parse, it gets checked against older pages for overlap and internal linking, it gets distributed to the channels where the audience actually is, and its performance gets tracked at the individual-page level, not folded into a site-wide number.

Stage 1 — Written to be found, not just written

New content should open with a direct answer in the first few sentences, use a heading structure that matches how people actually ask questions, and cite real sources — not just fill a word count. This is the difference between a post that reads fine and one that shows up when someone (or an AI answer engine) is actually looking for the answer it contains.

Stage 2 — Checked against what already exists

A single new post rarely moves the needle alone. The bigger opportunity on most firm sites is usually sitting in old pages that already rank but are thin, outdated, or badly structured — rewriting those often outperforms writing something new from scratch. A new post should link to those existing pages where relevant, not sit as an island.

Stage 3 — Distributed where the audience is

A blog post that only lives on the blog reaches whoever happens to find it organically, which is slow. The same idea, written natively for LinkedIn, Instagram, and other channels — not the same paragraph copy-pasted — extends its reach without diluting it.

Stage 4 — Watched, not filed away

This is the stage most firms skip entirely. Per-page traffic, keyword movement, and click data should be checked regularly, because that’s what tells you whether a piece is working or needs a rewrite. Publishing without tracking is publishing blind.

FAQ

Does a law firm need to write new content constantly to see results?
Not necessarily. Rewriting existing pages that already have some traffic or ranking is often faster to show movement than starting from zero on a brand-new topic.

How long before a new post shows measurable traffic?
It varies by competition and topic, but per-page tracking should start from day one so you have a baseline, not a guess, when results do show up.

What’s the minimum a firm should check monthly on its own content?
Which individual pages are getting traffic and how their rankings are trending — not just a site-wide total.

Any firm who’s worked with an agency knows that Thirty pages of activity metrics is not an answer. It’s a document designed to feel thorough while avoiding the one question that actually matters: what’s working, specifically, and what should stop.

Quick answer

A long marketing report padded with impressions, reach, and engagement percentages usually exists to demonstrate effort, not results. The metrics that actually tell a firm whether its marketing is working are narrower and more specific: per-page traffic, keyword ranking movement, and clicks — tracked individually, not rolled into a site-wide total.

Why the padding happens

Site-wide totals always go up over time as a site accumulates content, regardless of whether any individual piece is doing well. That makes a big rolled-up number an easy thing to point to, even when the underlying picture is mixed. A 30-page report with charts trending upward can be entirely consistent with half the firm’s content doing nothing.

What a useful report actually looks like

• Page-level, not site-level. Which specific pages are ranking, for which terms, and how that’s moved since last month.
• Comparable over time. The same pages tracked month over month, not a rotating set of vanity metrics.
• Short. If it takes 30 pages to say what’s working, something is being obscured, not explained.

What this doesn’t solve

Being able to see which pages are getting found and clicked is not the same as knowing which one closed a client. That link — a specific asset to a specific signed matter — requires the firm to confirm it, because it lives in a CRM or intake system the marketing platform doesn’t have access to. Anyone claiming to trace that automatically is skipping past a step that doesn’t exist yet in most law firm marketing stacks.

The one thing worth asking your agency this month

Skip the report. Ask for a list: every page on the site, its traffic this month versus last month, and its current search ranking for its target term. If that list can be produced in five minutes, the agency already has the data — the 30 pages were optional. If it can’t be produced at all, that’s worth knowing too.

FAQ

Why do agencies send such long reports if the real data is simpler?
A long report with rising totals is easier to present as progress than a short one that shows some pages up and others flat or down — even when the short one is more accurate.

Is a rising site-wide traffic number a bad sign?
Not necessarily, but a lot can be missed. A handful of strong pages can carry a rising total while most of the site does nothing. Page-level data is what tells the two apart.

What should replace the monthly report format entirely?
A live, per-page view a firm can check anytime, rather than a document assembled once a month.