AI Overviews Cut Clicks by a Third on Informational Searches. Where Do Law Firm Leads Come From Now?

Short answer: Two studies this year put numbers on what firms have seen in their analytics. Ahrefs (April 2025, 300,000 keywords) found the top organic result loses about 34.5 percent of its clicks when an AI Overview appears. Pew Research (July 2025, tracked browsing of 900 adults) found people clicked a traditional result on 8 percent of searches that showed an AI summary, against 15 percent without one, and clicked a link inside the summary on 1 percent. Both studies are dominated by informational queries. Law firm leads come mostly from transactional queries ("divorce lawyer Pasadena"), where the map pack still leads and Overviews are rare. Traffic reports will fall. Calls should not, if the site was built for the calls rather than the visits.

What did the studies find?

Both are consistent with what Search Console shows on firm sites we manage: impressions flat or up, clicks down on the informational pages, clicks steady on the practice and location pages.

Which legal queries get an Overview?

From our monthly Los Angeles query list (observation, not a study): process and definition questions almost always; cost questions sometimes; "lawyer near me" and "[practice] attorney [city]" rarely. The queries that produce calls are in the last group. The queries that produce blog traffic are in the first.

So a firm whose report is mostly blog traffic sees the drop. A firm whose report is mostly practice-page and map-pack traffic sees less of it.

Where do leads come from now?

The same places they came from in 2023, with a shift in weight:

1. The map pack. Still the largest single source of first calls for local practice queries. Unaffected by Overviews in our sample.
2. Practice and matter pages ranking for transactional queries. Steady.
3. Directory profiles (Avvo, Justia, FindLaw). Steady to rising, because AI answers cite them.
4. Referrals from content that was read but not clicked. New. A person reads an AI summary that cites a firm, does not click, and later searches the firm by name. This shows up as branded search and direct traffic, not as a click from the article.
5. Informational blog traffic. Down, and it was never the lead source it was reported to be.

The fourth item is the one to understand. Being the cited source in an answer the reader did not click is still worth something; the brand search a week later is the evidence. It just does not appear in the article's traffic line.

What should a firm measure instead of visits?

• Calls and form submissions, with source. Call tracking on the website number, not on the directory listings.
• Branded search impressions in Search Console, month over month. This is where uncredited citations show up.
• Map-pack position for the core query in each target city, weekly.
• Which of the firm's questions carry an Overview, and who is cited. If a directory is cited for a question the firm answers better, the page needs the answer at the top.
• Practice-page impressions and clicks, separately from the blog.

A report that leads with "sessions" is measuring the thing that is falling for reasons unrelated to the firm.

What should a firm change on the site?

Little, if the pages were built for calls. If they were built for traffic, three things:

1. Put the 40-to-60-word answer under each question heading, so that when an Overview quotes the page it quotes the firm by name.
2. Make sure every informational page links to the matter page and the phone number, so the reader who does click has a next step.
3. Stop producing informational posts with no path to a matter. Their traffic is what the studies say is gone.

What does not work?

• Blocking Google from the blog to "stop it being summarized". Removes the pages from the results entirely and does nothing about the Overview, which uses other sources.
• Reporting the traffic drop as a ranking problem. The rankings are usually unchanged. The Overview sits above them.
• Buying traffic to replace it. Paid clicks in Los Angeles legal are among the most expensive in the country; replacing free informational visits with paid transactional ones is a different business decision.

How does this change pricing?

It changes what a retainer should be judged on. A firm paying for SEO should ask for calls, map-pack position, practice-page performance and citation share, not sessions. Our own reporting was rebuilt around those this spring; the numbers are in the pricing page. To see the firm's current split between informational and transactional traffic and which Overviews cite it, start with the free visibility check. The wider argument is in the SEO vs GEO comparison.