USA Today, Reuters and Politico Weigh Cutting Off Google as Google Zero Arrives
Nilay Patel coined “Google Zero” in 2024 to describe a hypothetical: the moment search stops sending traffic anywhere except back into itself. It was a thought experiment then. It’s now a line item in strategy decks at some of the largest newsrooms in the United States.
USA Today has been weighing whether to block Google outright. Reddit is reportedly reassessing a deal worth around sixty million dollars a year after watching its traffic fall anyway. Politico and Reuters have both looked at limiting what Google can take. None of them has pulled the trigger, and the hesitation tells you as much as the consideration does.
Why this was unthinkable and now isn’t
The math used to be simple. Search sent between a fifth and two fifths of external traffic to most large publishers, making it the single biggest referral source by a wide margin. You do not switch off your biggest referral source. Any executive who proposed it would have been asked to explain themselves in front of a revenue chart.
What changed is that the chart stopped defending Google. Semrush data covering June 2025 to June 2026 shows traffic falling more than forty percent at some publications over that single year. Not a soft decline, not a seasonal dip, and not something a content refresh fixes. When the referral source is shrinking that fast on its own, refusing it costs less every quarter.
The crawler problem sharpens it further. The same bots that index a site for search also collect material that ends up training and grounding AI answers. Publishers can turn the bots away, but doing so historically meant vanishing from the results. That was the whole trap. The calculation only works when the results are worth less than the content being handed over, and for a growing number of titles that crossover has already happened.
Not everyone is losing
The Guardian and the BBC both gained traffic over the same period, which complicates the tidy narrative and is worth sitting with. Neither is a paywalled American metro daily competing on commodity news. Both have enormous direct audiences, strong app usage, global brand recall and a habit among readers of typing the name rather than searching a topic.
That’s the pattern in the exceptions. Sites people go to survive. Sites people arrive at do not. AI answers intercept the arrival. They can’t intercept a bookmark, a push notification, a podcast feed or an email that lands at seven in the morning.
Which means the strategic response to Google Zero isn’t really about Google. It’s about whether a publisher ever built a relationship with its audience or simply rented attention from an intermediary and called the rent a business model.
What blocking would actually cost
A publisher who blocks loses three things at once: the residual search traffic, the Discover traffic that often exceeds it, and the negotiating position that comes from being inside the system rather than shouting at it from outside.
They gain leverage only if enough others do the same at the same time, and there’s no mechanism for that. Coordinated withdrawal by competing commercial publishers raises antitrust questions in most jurisdictions. Individual withdrawal is a unilateral disarmament that Google barely notices. This is a collective action problem wearing a press freedom costume.
The realistic path is narrower and less dramatic. Selective blocking of the AI-grounding crawlers while staying indexed for classic search, if the technical separation ever becomes reliable. Licensing deals that put a number on the content. Aggressive investment in the direct channels that the answer engines can’t stand between.
None of that is a stand. It’s triage, and the outlets doing it early will be in better shape than the ones still waiting for the traffic to come back.
It isn’t coming back.