Hope Sterling: Juniper, hey — okay I've been staring at this story since yesterday and I literally cannot move on until we talk about it.
Juniper Vale: The Wall Street Journal thing? Yeah, me too — something about the Newsweek number just keeps sitting wrong with me.
Hope Sterling: RIGHT — because The Wall Street Journal dropped this on July 21st and it's like, okay, USA Today, Politico, Reuters, The Economist, People Inc., Reddit — they're all apparently sitting around going 'should we just... block Google?' And I'm reading this like, is that even a real threat?
Juniper Vale: That's what we're trying to work out today, honestly — whether any of this is a real exit or just the loudest negotiation we've ever seen.
Hope Sterling: And the thing that like — okay, this is the part that broke my brain — Newsweek. Their organic search traffic went from 10.16 million monthly visits in November 2025 to basically zero by summer 2026. Zero! And they expect a hundred million dollars in revenue this year, which is their highest since 2012. Because AI companies are paying them directly now.
Juniper Vale: That's the disorienting part. Traffic collapses and revenue goes up — and Google still controls roughly 88% of referral traffic for most publishers. So you've got this situation where the people threatening to block their biggest source of visitors are somehow... fine?
Hope Sterling: Which is why Mike Reed saying 'it's time to take a stand and say enough is enough' hits so different when you put those numbers next to each other.
Juniper Vale: And Neil Vogel at People Inc. saying blocking is a hundred percent on the table — I mean, does he know something Newsweek figured out first, or is he just the most willing to say the quiet part loud?
Hope Sterling: But okay — like, that's the part I need someone to actually explain to me, because I don't understand why it broke so fast. Like what actually changed?
Juniper Vale: Think of it this way. Google used to be a waiter who handed you a menu and pointed you to the restaurant. Now it just reads the menu to you at the door and you never go inside. That's it. That's AI Overviews.
Hope Sterling: Okay that — yeah, that actually clicked for me.
Juniper Vale: And the timing is — I mean, this is where I keep getting stuck — Google rolled out AI Overviews to every US user in May 2024, then expanded globally by May 2025. That's thirteen months. And the Cloudflare data shows AI-training crawlers went from 22% of all crawler traffic in spring 2025 to 52% by June 2026. That's not gradual. That's a flip.
Hope Sterling: Wait — 52%? More than half of all crawler traffic is now AI training bots? Not actual readers?
Juniper Vale: More than half. And the human side of that equation is falling — Cloudflare's own data, cited by The New York Times, shows human traffic to software, retail, finance, IT — down as much as 40% in under a year. The crawlers are going up, the people are going away.
Hope Sterling: Which is why the Semrush numbers are so — like, Business Insider down more than 85%? USA Today cut in half? Politico down 23%, CNN down 25%? All in the year ending June 2026? That's not a bad quarter, that's a structural collapse.
Juniper Vale: And that's actually why 'Google Zero' landed as a term — publishers aren't using it to describe where they are, they're stress-testing against it as a planning scenario. Like, what does our business look like if referrals hit zero. The fact that they're running that model at all tells you how close it feels.
Hope Sterling: So the thing that's actually new isn't that Google is powerful — everyone knew that — it's that the deal that made it livable, you send traffic, we give you content, that deal just... broke. And nobody signed off on ending it.
Juniper Vale: But that's exactly where the circulating take falls apart — because everyone's saying 'publishers are finally holding Google accountable,' and I want to know who actually believes that. Because the block isn't a weapon. Googlebot can't be split. If you block it, you don't just stop AI Overviews — you disappear from Google Search entirely.
Hope Sterling: Wait — like, all or nothing? There's no middle setting?
Juniper Vale: All or nothing. And 88% of your referral traffic just walked out the door with it. That's not fighting back, that's — I mean, that's mutually assured destruction, except the destruction lands almost entirely on the publisher.
Hope Sterling: Okay but — wait, what about the opt-out tool? Google-Extended? Because I feel like I've seen that floated as the thing publishers are supposed to use, like 'oh just opt out'—
Juniper Vale: Google-Extended only blocks content from training Gemini. That's it. It does nothing for AI Overviews, which is actually the thing eating the traffic. Publishers described it as functionally useless for their core concern. It's a fig leaf.
Hope Sterling: Stop. So the tool Google gave publishers to 'opt out' doesn't even touch the actual problem? Like — okay, that is actually insane. Because here's what that looks like in real life: a USA Today reporter publishes a breaking scoop at 6 AM, and by 6:15, Google's AI Overview has already summarized the whole thing at the top of the results page. Reader gets the answer. USA Today gets literally zero traffic. And the opt-out tool? Does nothing about that.
Juniper Vale: Nothing. And that's the scenario Mike Reed is reacting to when he says 'enough is enough.' Not some abstract future — that's happening at 6:15 every morning.
Hope Sterling: So if blocking is self-destruction and the opt-out is a decoy, what's actually left? Because — and I'm gonna come back to this — Reddit got sixty million dollars a year from Google and is STILL reportedly thinking about walking at renewal, which tells me the money alone isn't solving whatever this actually is.
Juniper Vale: The only body right now seriously consulting on the technical fix — crawler separation, so publishers could actually block AI training without nuking their search presence — is the UK Competition and Markets Authority. That's what their consultation opened on. And the fact that it takes a regulator to even attempt it tells you publishers genuinely cannot solve this unilaterally. The real 'fighting back' isn't a block. It's waiting on that process.
Hope Sterling: But that's the part that's making my head spin — like, Reddit already ran that experiment. Sixty million dollars a year, structured access, the whole thing. And they're still sitting there in 2026 going 'hmm, maybe we block anyway'? That's not a satisfied customer.
Juniper Vale: Which is — I mean, that's the signal, right? The compensation model exists. Reddit proved it can exist. But Reddit is still evaluating an exit, which means the money didn't actually fix what was broken. And what was broken is the power question, not the revenue question.
Hope Sterling: Wait, so what IS the power question, specifically?
Juniper Vale: Who decides how your content gets used after you publish it. That's it. Reddit got paid — but Google still determines whether Reddit's answers show up in AI Overviews, in what form, summarized how. The check doesn't change that. And Newsweek's pivot is actually the cleaner answer — stop being a traffic business entirely, sell the data directly to AI companies, hit a hundred million in revenue, fifteen-year high. You exit the Google dependency loop altogether.
Hope Sterling: Okay but — and I say this as someone who reads Newsweek sometimes — if publishers become data licensing businesses, what happens to the actual journalism? Like, who's paying a reporter to go sit outside a courthouse for six hours?
Juniper Vale: That's — yeah, that's the real watch item. Because the CMA's crawler separation proposal is the one thing that could give publishers the technical middle ground — block AI training without disappearing from search. But it's UK-only, it's unresolved, and the jurisdiction stops at the border. What matters is the precedent. If the CMA mandates that separation, every licensing negotiation globally gets that ruling cited in it.
Hope Sterling: So the UK ruling becomes like — the template everyone else waves around at the negotiating table.
Juniper Vale: Exactly that. And Google has added guardrails — news and health query restrictions in AI Overviews — but publishers say it hasn't moved the needle on actual click-through. So we're watching a moment where the structural fix is sitting in a UK consultation room, unresolved, and the only publishers who've found revenue stability did it by walking away from the traffic model entirely. That's what to watch: does the CMA rule, and does Reddit walk.
Hope Sterling: The whole question comes down to this — does Reddit walk at renewal, or does the sixty million dollars just... win? Money versus power.
Juniper Vale: And honestly? I don't think anyone knows. Not Reddit, not the CMA, not Google. That's — I mean, that's the actual state of it.
Hope Sterling: Which is kind of wild to just sit with.
Juniper Vale: Yeah. Thanks for working through it with me — I feel only slightly less confused.
