Traffic Is Down But Conversion Quality Is Up

Key Points

• Brainlabs data covering 54 clients shows organic search sessions fell 10.5% after Google AI Overviews hit 30% search result penetration in September 2025.
• AI referral traffic rose 163% in the same period, but users arriving via ChatGPT, Gemini, and Copilot convert at 1.5x the rate of organic search visitors.
• The traffic that's disappearing is mostly casual, top-of-funnel. The traffic staying is higher-intent.
• Publishers and advertisers need to fold AI referral channels into standard measurement KPIs, not treat them as experimental noise.
• Less raw traffic doesn't automatically mean less revenue. It does mean your monetization setup needs to work harder per session.

What Happened

Digiday's August 2026 analysis published data from Brainlabs, a marketing services agency that tracked Google Analytics across 54 clients over 14 months. Once Google AI Overviews started appearing on at least 30% of search results pages in September 2025, organic traffic fell sharply for most of those clients.

Overall sessions dropped from 140.1 million to 125.4 million, a 10.5% decline. Forty-six of 54 clients saw a dip. Ahrefs has reported AI Overviews cutting traffic to some sites by as much as 58%.

The Conversion Data Changes the Story

Sam Paez, senior SEO manager at Brainlabs, noted that AI Overviews are primarily triggering against informational and educational queries, not transactional ones. The traffic that's disappearing tends to be users who wanted a quick answer and never had serious purchase intent anyway.

The users clicking through from AI platforms are different. Brainlabs tracked "key events" alongside sessions, measuring actions like purchases, newsletter signups, and page completions. AI-driven key events rose 335% during the measurement window. ChatGPT, Copilot, Gemini, and Perplexity referrals converted at 1.5x the rate of traditional organic search traffic.

That's a meaningful signal about audience quality.

Essential Background Reading:

• AI and Publishers Resource Center: Everything publishers need to understand about AI's impact on traffic, monetization, and strategy
• What Publishers Need to Know About AI Bot Traffic in 2026: A practical guide to understanding AI bot traffic, ad quality implications, and what to do about it
• AI Bot Traffic Is Now 40% of the Web: Why publishers need to pay close attention to the scale of AI crawling activity across the web

The Sector Breakdown Publishers Should Care About

The impact isn't uniform. Brainlabs' dataset covered 19 sectors and 16 retailers, and the variance is significant.

Retail held up best because AI Overviews don't trigger as often against transactional searches. Categories where consumers research heavily before buying took the hardest hit on raw traffic volume.

For publishers in entertainment, sports, or gaming verticals, this data points to a more resilient position than the doom headlines imply.

Related Content:

• Google's AI Search Updates Give Publishers a Real Shot at More Clicks: How recent Google changes create new click-through opportunities for publishers willing to adapt
• What Publishers Actually Gain from Google's AI Search Updates: A direct look at the publisher-side benefits buried inside Google's AI search rollout
• Yahoo Scout Shows AI Search Doesn't Have to Cannibalize Publisher Traffic: An alternative model for how AI search can send traffic to publishers rather than replacing it
• Agency Revenue Drops Signal AI Search Traffic Shift: Industry-level data showing how the AI search transition is already affecting publisher revenue downstream

What This Means for Publisher Revenue

The analysis gets practical fast for publishers running programmatic monetization.

Raw session volume dropping 10% doesn't translate to a 10% revenue drop, but only if your monetization is optimized at the session level. Publishers whose RPS is lagging are going to feel this harder. Publishers whose ad stack is tuned to extract maximum value from each session will absorb the hit better.

There are also measurement gaps Paez flagged directly: some AI referral visits are recorded as direct traffic, because users open a new browser window rather than clicking through inside the AI interface. That means AI referral numbers in most analytics setups are likely undercounted. Publishers looking at their traffic reports may be underestimating how much of their audience is already arriving via AI channels.

A few things publishers should be doing now:

• AI referral traffic attribution: audit your analytics setup to capture AI-sourced visits currently logging as direct. Segment and track them separately.
• Session-level monetization: if your RPS is optimized, a smaller audience of higher-intent users can yield comparable or stronger revenue than a larger, lower-intent one.
• Key event tracking over session volume: shift internal reporting to weight conversion events alongside traffic. Raw sessions are increasingly a vanity metric.
• Vertical context: if you're in retail, beauty, or entertainment, the data suggests you're better positioned than average. Finance and insurance publishers should prepare for continued pressure on organic volume.

Next Steps:

• AI Traffic Is the New SEO: How publishers can actively optimize their content and structure to capture more AI referral traffic
• Google AI Mode Is Here: Your Traffic Strategy Needs to Catch Up: Practical steps for updating your traffic acquisition strategy in response to Google's AI Mode rollout
• Are AI-Generated Summaries Reliably Driving Publisher Traffic: Data on whether AI platform citations actually translate into measurable referral sessions for publishers
• The Quality vs. Quantity Revolution: Why ad load and traffic stability matter more than raw volume in a session-level revenue environment

The Measurement Problem

AI referral traffic needs to be in your everyday measurement KPIs, not in a separate "emerging channels" bucket reviewed quarterly.

ChatGPT hit 1 billion monthly active users in May 2026, per Sensor Tower data cited in the Digiday piece. Claude had 56 million MAUs in the same period. The volume is real, the conversion quality is real, and the attribution gaps are real.

Publishers who treat this as a future problem are already behind. The session data is being recorded now. The conversion events are happening now. The question is whether your reporting setup can see them.

See It In Action:

• AI Crawler Impact on Lifestyle Publisher Traffic: Real ecosystem data on how AI crawlers are affecting traffic and ad revenue for lifestyle publishers specifically
• How AI Crawlers Impact Entertainment Website Traffic and Ad Revenue: A vertical-specific look at AI crawler effects on entertainment publishers' traffic and monetization
• Brands Measure AI Traffic Customer Journey Impact: How brands are beginning to quantify AI-sourced traffic within the customer journey, with implications for publisher monetization

Making the Traffic You Have Work Harder

Traffic volume shifting is a structural change, not a temporary blip. The publishers who come out ahead will be the ones whose monetization extracts more value per session, not the ones waiting for organic volume to recover.

Our RAMP platform is built around exactly this problem: optimizing yield at the session level, across every ad format and demand source, so that fewer sessions don't automatically mean less revenue. If your current setup scales linearly with traffic and traffic is falling, that's a yield ops problem with a fixable answer.

The AI search transition is compressing the top of the funnel. Make sure everything below it is running as efficiently as possible.