What’s really behind the drop in search traffic to news publishers?
Product Initiative Newsletter Blog | 24 June 2025
Hi there.
You may have seen my colleague Greg Piechota’s LinkedIn note about this news headline: “Chatbots are replacing Google’s traditional search, devastating traffic for some publishers.” We went down quite the rabbit hole as a team. Today a share some views and research about that, plus a glimpse into Google’s new AI Mode.
By the way, are you part of the conversation on INMA’s Slack? Hope to see you on the #product-tech-initiative channel soon!
Hope to see you somewhere online or in person soon.
Jodie
What’s really behind the drop in search traffic to news publishers?
Over the past few weeks, there’s been a flurry of discussion among INMA’s initiative leads about the steep decline in search traffic to news publishers — particularly in the U.S. While headlines have pointed to AI tools like Google’s AI Overviews as the cause of this disruption, the reality, as usual, is more nuanced.
Reports of search referral drops in excess of 50% have caused alarm, with some publishers pointing to AI as the primary culprit. But Grzegorz Piechota, INMA’s Researcher-in-Residence, pushed back on that narrative. Drawing from multiple data sources, he argued that while AI is a contributing factor, it’s only part of a broader set of changes.
To get a clearer picture, we dug deeper. I used ChatGPT to help collate and visualise data. The chart below maps out what’s really happening.
Connecting the dots
We know AI is changing the shape of search, but it hasn’t yet scaled to the level that would explain such large traffic losses on its own. Google’s AI Overviews only appear on a small fraction of news-related queries, and the more immersive AI Mode is still being tested in the U.S.
Instead, we looked at all the contributing factors that could explain the drop in referrals from 2022 to 2025. Here’s what the data shows:
While AI-related experiences are a growing concern, much of the damage to search referrals predates their rollout. Algorithm changes, volatility in Google Discover, the collapse of commerce SEO strategies, and broader behavioural shifts have all played a role.
There is also a large geographical point to note. It’s striking how regionalised this disruption appears to be — at least for now. The U.S. is clearly the test market for new AI features from Google, and it’s where the steepest declines are being reported. Some publishers there are seeing real impact from AI Overviews and Mode, particularly in query areas where those tools suppress click-throughs.
If we compare this to Europe, we see that European publishers have yet to feel the full effect. AI adoption in search remains low across the continent, with Google continuing a slow and cautious rollout. But that doesn’t mean Europe is insulated — only that it may be six to 12 months behind the U.S. curve.
After a number of prompts, here’s where ChatGPT landed:
🌎 U.S. vs Europe: Regional AI Impact
🔹 United States
AI Overviews presence: 3.47% of trending U.S. news queries trigger AIO — the highest globally.
Recent decline in organic search: Since introducing AIO in late 2023 and AI Mode rollout (March-May 2025), top U.S. publishers have seen sharp drops: HuffPost, The Washington Post, Business Insider all experienced >50% reduction in Google search referrals since April 2022.
Zero-click trend: 59% of Google searches in the U.S. end without clicks to external sites, well above pre-AI levels.
🔹 Europe
Lower AIO penetration: UK at ~1.30%, Germany at ~0.95%; average EU rate ~0.3%–1.3% .
Zero-click slightly higher but still lower external referrals: ~37.4% open-Web clicks per 1,000 EU searches, vs. 36% in the U.S., indicating fewer link clicks overall.
Organic search traffic decline: Publishers in the EU show less pronounced losses — The New York Times’ EU SEO fell ~3-5% YoY while U.S. rates are much steeper (~20%-50%).
So, what should we take from this?
It is not a single-source problem. It’s a confluence of forces, some short-term, others structural. AI is the headline story now — and rightly so, as it represents a long-term shift in how people discover and consume information. But it’s not solely to blame for recent declines.
Here’s where we landed as a group:
Don’t oversimplify the problem. AI is accelerating change, but the groundwork was laid long before AI Overviews.
Segment your traffic sources. Understand the difference between Discover, traditional Search, and other Google surfaces. Not all are declining — some are simply reshuffling.
Watch regional rollouts. What’s happening in the U.S. is a preview of what’s likely to come elsewhere.
Prepare for an AI-first search environment. Even if the numbers are small today, the underlying direction of travel is clear.
This is the time for strategic clarity. We need informed, coordinated action across product, audience, and editorial teams.
So while AI is the current and most strategic concern, it’s not yet the largest factor behind historical traffic drops. The threat going forward is qualitatively different: it’s systemic and structural, not purely algorithmic. That’s why it feels, and likely is, more existential.
Date for the diary: July 2, INMA Webinar on how AI platforms are reshaping news
In this INMA Webinar, Reuters Digital News Report 2025: How AI and Platforms are Reshaping the News Ecosystem, Nic Newman presents the latest global data, followed by a discussion with Lucy Kueng and me on what it all means for publishers.
From AI’s impact on discovery to rethinking access models, this session offers strategic insights for navigating a post-search, post-social news landscape. It’s free to members, and you can sign up here.
Google AI Mode is here and this is what it looks like
I was mid-Zoom with some colleagues, searching for something related to our conversation, when Google popped up with a prompt: “Would you like to try AI Mode?”
There was no question, of course. I clicked yes.
We’ve been talking about the evolving nature of search and the impact Google AIO may have on traffic to news sites. But AI Mode takes this a step further — and it’s now clear to me just how dramatic that shift could be.
Back in March, a senior search executive told me news publishers should be planning for a “near-zero search strategy” within two years (more on this from INMA’s April report As Search Ends for News, Here is What’s Next). At the time, it sounded provocative. Now, it feels inevitable.
The phrase “Google Zero” is cropping up more often. Check out Nick Thompson on search and Neil Vogel on Google Zero. I’ll dive into this more in the coming months.
As you’d expect, the AI Mode interface is clean and intuitive. It invites users to ask “detailed questions,” nudging them away from the keyword-based habits of traditional search.

Suggested prompts help guide the experience, and the layout mirrors other answer engines we’re starting to see across the ecosystem.
When I tested it with a colleague who was German, we tried a few contentious questions. Some returned traditional results — albeit with a slightly different design — while others triggered the full AI Mode experience: conversational answers, no links below, and a seamless prompt to follow up with another question. It’s AIO but without the fallback to classic search below it.
Sources are cited subtly — small link icons appear inline, unbranded, but clickable. A note in the top right corner shows how many sources were used with three clearly listed.

Interestingly, we’ve heard from several publishers that conversations are underway about being featured in those top-right source boxes — but only if they have a Gemini or Cloud partnership. It seems AI companies are beginning to lock publishers into their ecosystem.
Times are changing, fast. As noted in the post above, some countries and companies may not have seen impact yet. But I think we can all agree that they will soon. Now is the time to prepare for an AI answer engine-first world.
About this newsletter
Today’s newsletter is written by Jodie Hopperton, based in Los Angeles and lead for the INMA Product and Tech Initiative. Jodie will share research, case studies, and thought leadership on the topic of global news media product.
This newsletter is a public face of the Product and Tech Initiative by INMA, outlined here. E-mail Jodie at jodie.hopperton@inma.org with thoughts, suggestions, and questions. Sign up to our Slack channel.