Hi everyone, this is Neo.

Today I want to talk about something that left me with mixed feelings — Google’s “official explanation” for the traffic collapse hitting independent sites and media publishers.

If your independent site has watched its Google search traffic slide for the past year, you’ve probably asked yourself: Did I do something wrong, or is Google’s AI Overviews siphoning off my traffic?

Over the past six months, Google’s head of search, Liz Reid, has given independent sites and publishers a “reassuring” answer in three separate settings — once on Google’s official blog, once with the Wall Street Journal, and most recently on Bloomberg’s Odd Lots podcast:

“Relax — AI Overviews only eliminated ‘bounce clicks,’ the low-value visits from people who click in and immediately bounce back. Users who actually want to read deeply will still click through to your site.”

In other words: “What you lost was junk traffic anyway. Nothing to mourn.”

Sounds reasonable, right? But here’s the problem — Reid has never produced a single piece of data to back this up.

And independent third-party data tells a completely different story.

Today I’m going to lay out the whole saga: what happened, what the data really says, and what it actually means for independent sites.

Liz Reid’s “bounce clicks” theory

First, Reid’s core argument. On the April 23, 2026 episode of Bloomberg’s Odd Lots podcast, hosts Joe Weisenthal and Tracy Alloway asked her directly: what impact are AI Overviews having on publisher traffic and ad revenue?

Reid’s answer can be boiled down to three points:

1. AI Overviews only eliminated “bounce clicks”

People who used to click in and immediately head back to the search results no longer need to visit your page — they get the answer straight from the Overview.

2. Readers who genuinely want depth still click through

People looking for longer, deeper content will still click into your site.

3. Ad clicks are down, but rising query volume makes up for it

Ad clicks on some queries have dropped, but overall search volume is up, so it balances out.

It all sounds tightly reasoned and internally consistent.

But this narrative has one fatal flaw: Reid has never offered any data to support it.

What Reid didn’t say

Let’s look closely at Reid’s three public statements over the past eight months:

1. Google’s official blog, August 2025 She said organic clicks from Google Search to websites were “relatively stable year over year,” and that “high-quality clicks” — visits where users don’t quickly bounce back — were growing.

What was in the blog post?

• ✅ A claim
• ❌ No charts
• ❌ No specific percentages
• ❌ No year-over-year comparison data

2. Wall Street Journal interview, October 2025 She explicitly used the term “bounced clicks,” saying ad revenue with AI Overviews was “relatively stable.”

What was in the interview?

• ✅ Another claim
• ❌ Still no data

3. Bloomberg’s Odd Lots podcast, April 2026 She told the hosts that Google tracks whether users “return to search more frequently” as one of its key signals.

What was in the podcast?

• ✅ Another claim
• ❌ Still no concrete numbers
• ❌ And the hosts never pushed her for evidence

Neo’s take: as a content operator, here’s the pattern that should put you on alert: a platform that controls all the data offers an explanation for a widely questioned phenomenon — yet refuses to publish any data the outside world can verify.

This isn’t conspiracy theory. It’s a basic issue of scientific method — an explanation with no data behind it isn’t a conclusion, it’s a claim.

What independent third-party data actually shows

Fortunately, even though Google won’t publish its numbers, independent research firms have been tracking what’s really happening.

And their findings almost completely contradict Reid’s “low-value clicks” theory.

1. Chartbeat / Reuters Institute for the Study of Journalism

Google search traffic to publishers worldwide has fallen by about a third (roughly 33%).

That’s Chartbeat data cited in the Reuters Institute’s 2026 Journalism, Media, and Technology Trends and Predictions report, covering more than 2,500 publisher sites.

And it gets worse:

Google Discover referral traffic is down 21% year over year.

If the “bounce clicks” theory held — that what was lost was only low-value, quick in-and-out traffic — then why did even Google Discover (a recommendation feed, where users arrive specifically to discover new content) drop 21%?

2. Seer Interactive

Organic CTR on queries with AI Overviews fell from 1.76% to 0.61% — a 61% drop.

That’s the 2024-to-2025 comparison.

Seer added a crucial nuance: most of these are informational queries, which historically already have low CTRs. In other words, a CTR that was already low dropped another 61%.

3. Pew Research Center

After studying 68,000 real search queries, Pew found:

• When AI Overviews appeared, only 8% of users clicked a search result
• Without AI Overviews, the click rate was 15%

The click-through rate was nearly cut in half.

4. Digital Content Next (members include The New York Times, Condé Nast, Vox, and others)

Data from its 19 member publishers in May–June 2025 showed:

Median year-over-year Google search referral traffic down 10%.

DCN CEO Jason Kint said at the time that the member data provided “ground truth” about what was happening to publisher traffic.

Neo’s take: put these four independent datasets together and you get:

• Overall traffic down 33%
• CTR on informational queries down 61%
• Click-through rate cut in half when AI Overviews appear
• Referral traffic to premium publishers down 10%

Every single dataset carries far more weight than Reid’s “we just eliminated bounce clicks.” If it were only low-value traffic, how could overall traffic to publishers worldwide fall by a third?

The biggest data gap of all

Reid’s “bounce clicks” theory and the independent data aren’t even measuring the same thing:

| Dimension | Liz Reid’s “bounce clicks” theory | Independent third-party data |