Cited pages get twice the clicks per impression, but still 38% fewer than before AI Overviews. A conversion model for fewer, warmer visitors.

By the Search Converts team · October 19, 2026

The click data, plainly

Seer Interactive tracked Google Search Console data across 53 brands from January 2025 through early 2026, 5.47 million queries. For informational queries on desktop:

Two conclusions sit inside those three numbers. Once the Overview exists, which you do not control, being cited is worth 120% more clicks per impression than being ignored. And even cited, you have lost 38% of the clicks the same impression produced before. Ahrefs’ February 2026 study across 300,000 keywords put the top-result drop at 58% over two years, and SparkToro’s analysis of Similarweb data found 68% of US Google searches ending without any click in early 2026.

This is the environment. Nobody is optimizing their way back to 2023 traffic. The work is to make each remaining visit count.

Why the remaining visitor is worth more

The person who read the AI summary, saw your name, and clicked anyway is not browsing. They want something the summary did not give them: a price, a booking, a human, a specific detail. The early data agrees. HubSpot, citing Similarweb, reports AI-referred visitors converting at 11.4% in ecommerce against 5.3% for standard organic traffic. Semrush values an AI-referred visitor at about 4.4 times an organic one. SE Ranking’s June 2026 study puts AI referrals at only 0.32% of all visits, so the segment is small, but it is the warmest traffic most sites receive.

The model

Run it on your own numbers; the structure is the point.

1. Start with impressions, not clicks. Search Console now reports AI Overview and AI Mode impressions separately (worldwide since August 31, 2026). Impressions are the stable quantity; clicks are what you are trying to recover.
2. Apply the three CTRs. Split your key queries into no-Overview, cited, and uncited. Use Seer’s figures as placeholders until you have 90 days of your own. The difference between the cited and uncited rows is the value of earning the citation; that is the search team’s job.
3. Apply two conversion rates, not one. Your historical organic conversion rate for the no-Overview and uncited rows, and a separate, measured rate for AI-referred visitors (segment by referrer: chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com, and the AI Mode and Overview referrers Google exposes). If you have not measured it yet, do not assume the 11.4%; assume your organic rate and let the data surprise you.
4. Multiply through to revenue. Visits by conversion rate by close rate by average value. The cited row with the AI conversion rate will often produce more revenue than the pre-AI row did with three times the clicks. That is the number that reframes the conversation with whoever is mourning the traffic.
5. Find the lever. Three variables are yours: the citation (search work), the conversion rate of the warm visitor (page work), and the close rate (follow-up). The one with the biggest gap to its benchmark is where the next dollar goes. Our conversion calculator runs the multiplication; the one-percent lift post shows why the page work usually wins.

What raises the warm visitor’s conversion rate

• Continuity. The first screen confirms what the answer said: same service, same range, same location, same proof. A visitor sent by a summary that quoted your price should not land on a slogan.
• One action, one tap. The warm visitor is ready. A phone number that dials, a booking button that books, a form short enough to finish on a phone. Zuko’s benchmarks put form completion near 52% once started and lower on mobile; every field is a tax on the visitor you worked hardest to get.
• Proof the assistant could not show. Photos of the actual work, the named technician, the review text, the license number. The summary gave a sentence; the page gives the reasons.
• A reply inside the hour. A warm lead cools at the speed of your follow-up. The speed-to-lead calculator prices the delay.

We package this as AI traffic conversion optimization for clients who want it run and measured; the model above is free to anyone with a spreadsheet.

A worked example

A service business with 10,000 monthly impressions on its key informational queries, a $1,200 average customer and a 30% close rate. Before Overviews, at 3.35% CTR and a 4% page conversion rate: 335 visits, 13.4 leads, 4 customers, $4,824. After Overviews, cited, at 2.07% CTR, with the same 4% conversion: 207 visits, 8.3 leads, 2.5 customers, $2,981. That is the loss everyone feels. Now measure the AI-referred segment and find it converting at 8% (a conservative version of the published figures): 207 visits, 16.6 leads, 5 customers, $5,962. Same impressions, fewer visits, more revenue. The page did that, not the traffic.

Sources: Seer Interactive, AI Overview impact on Google CTR, 2026 update (April 24, 2026); Ahrefs, AI Overviews reduce clicks, 2026 update; SparkToro / Similarweb, zero-click search in 2026; HubSpot, AEO trends (Similarweb conversion data); Semrush, AI SEO statistics; SE Ranking, AI traffic research study (June 2026); Google, new controls and reporting for website owners; Zuko, form benchmarks.