Publishers at the Digiday Publishing Summit in Miami are planning for a structural decline in Google search traffic rather than hoping for a rebound. In closed-door town hall sessions, publishing executives described AI licensing deals as opaque, uneven and hard to benchmark, even as they test ways to earn money from AI visibility.
Google Zero becomes the working assumption
The phrase “Google Zero” does not mean search disappears overnight. Executives expect search referrals to keep declining and say the mass-traffic programmatic model that supported much of digital publishing has lost steam. One participant said the days of making money on mass Google referrals are generally down and over.
Google Discover may still grow, but publishers view those users as less valuable than traditional search visitors. The strategic shift is toward direct relationships: subscriptions, memberships, events, sponsorships and premium advertising. Some publishers are already seeing direct reader revenue grow; The Guardian reported £126 million ($170 million) in direct reader payments, up 17% year over year, with 1.4 million paying digital supporters.
The Verge’s director of audience and subscriptions, Esther Cohen, summed up the new posture onstage: Google traffic is now extra, not something to plan or optimize around.
AI licensing deals are still a black box
Publishers said AI companies arrive with different contract structures—flat rates, grounding rates and other models—without clear definitions of what the metrics mean. Because large buyers prefer private deals, licensing marketplaces have been slow to develop. Several executives said this prevents them from knowing what a fair deal looks like, even among peers.
At the same time, Google has been quietly inviting publishers into its AI Contribution Pilot, which pays when content “significantly contributed” to a response in Gemini, AI Overviews or AI Mode. Participants can see limited data and report receiving some payments, though Google decides what counts as valuable.
What publishers are testing
Publishers are experimenting with several AI-era tactics:
- Optimizing articles for AI surfaces: one publisher reported a 40% increase in customer acquisition and traffic.
- Using Google Search Console’s AI performance data to identify URLs appearing in AI Overviews, which can reveal impression-only listings and very low click-through rates.
- Refreshing content around priority entities such as people, places and brands to lift visibility in ChatGPT, with one publisher seeing same-day impression increases.
- Licensing shopping content inside an onsite AI chatbot, though participants described the revenue as modest.
Reuters’ general manager for digital, Phil Andraos, shared a related revenue-protection tactic: its dynamic paywall uses machine learning to estimate the ad revenue a user could generate, then adjusts access decisions to reinforce both subscriptions and ad income.
Why this matters for marketing and media teams
This is not just a publisher story. Marketers that depend on search visibility, content partnerships or publisher inventory should prepare for a world where Google referrals are smaller and AI surfaces are harder to measure. The publishers’ response—diversify into direct revenue, demand clearer licensing terms and treat Google as incremental—applies to any content-led business.
The wider signals point the same way. Amazon is adding short-form news clips to Prime Video, OpenAI has partnered with Canadian local news network Village Media on an AI-powered community tool, and publishers’ paid search budgets have risen 274% in three years, with the top 100 publishers spending an estimated $113 million in July according to Similarweb data cited in the briefing. Meanwhile, Reach is cutting 160 editorial jobs and McClatchy has reduced staff by more than 90 unionized workers across 17 publications, showing how quickly traffic dependence becomes cost pressure.
Source: Digiday
