Publishers are still on the long ramp down in referrals from organic search, although I still maintain that "Google Zero" is more marketing than prophecy. One hoped-for outcome commonly discussed is that AI crawlers, including Google, will enter into content licensing agreements with publishers.
My worry is that publishers will like a world where Google pays for their content about as much as they like the world where Google stops sending them traffic. News can come from anywhere. When a buyer can get the same facts from twenty sellers, the buyer sets the price, and the price is close to zero. The lucky publishers will come away with pennies. The unlucky ones, mostly small, will come away with nothing. General interest publishers with undifferentiated coverage will fare the worst.
New Old Web is a blog published by Alley and Lede. We're researchers, strategists, designers, and developers who want to make the internet a fun place to live and work.
In September, Digiday reported that Google has been quietly scaling something it calls the AI contribution pilot. Participating publishers get a Search Console widget that shows what they earned that month when their content "significantly" contributed to an answer in AI Overviews, AI Mode or Gemini. Google calls the model "pay per value" and of course, Google itself decides the value. Participants in the pilot told both Digiday and AdExchanger that the widget shows a number but not how the number was calculated.
Then, two weeks later, The Information put some figures on it. Notably an early participant is earning more than $1 million a year. A later entrant has collected $50,000 to $60,000. Several small sites have earned less than $1,000 over several months, which for them works out to less than a tenth of one percent of their ad revenue. Payments move month to month without explanation, and opting out of the money does not opt you out of the product.
Out of context, those revenue numbers aren't particularly meaningful, but the spread across orders of magnitudes resembles the distribution of ad revenue in the heyday of digital advertising for publishers.
What decidedly does not resemble the ad-supported era of online publishing is the substitution problem created by LLM interfaces to news and information. If it trusts the source, a chatbot needs the facts just once, where an engaged reader might follow several different sources for similar content, and even for such drivel as "what time is the Super Bowl?" lots of publishers could get some amount of search traffic.
The pilot data already shows this. Publishers told The Information that niche subjects with devoted followings like anime and gaming earn more than general coverage. A general interest publisher in the market for facts to feed AI is a commodity farmer.
And Google has no reason to price generously. One publishing exec made the point to Digiday: a company that has freely crawled and indexed the web for twenty-five years has little incentive to start writing big checks for content it already has. David Buttle, who founded the publisher coalition SPUR, called the per-use payments less a meaningful exchange than a strategic hedge. What's more, once a publisher takes payments, Google can tell regulators that it compensates publishers for AI summaries, and the publisher has given up its best argument.
Licensing rewards scarcity and bargaining power, which most publishers do not have.
We've seen this before
Australia got licensing in 2021, with the force of law behind it. The News Media Bargaining Code moved roughly A$250 million a year from Google and Meta to Australian news businesses. It is the strongest version of what publishers are asking for, and its outcomes are the clearest warning about what they will get.
The deals were voluntary and confidential, so the platforms negotiated with whoever had leverage and skipped the rest. WARC's read early on was that the biggest publishers secured large secret deals while the publishers who most needed the money were left out, exactly as predicted. A parliamentary review three years later confirmed this: the code's primary criticism was its inconsistent outcomes, with legacy outlets receiving most of the funding, which worsened the competitive disadvantage of new, small, online-only and regional publishers. Eighteen small publishers eventually banded together as the Public Interest Publishers Alliance just to get a hearing.
Is there a silver lining for local?
If licensing primarily pays for scarcity and leverage, the question for any publisher is: Do I have either? I think local news has a real answer on both counts.
Local publishers share this with niche publishers like the gaming and anime publishers reportedly getting paid under the content licensing pilot. Local news is inherently scoped down, so the number of outlets writing about any given event is smaller — in some cases, too small, and scarcity gives rise to news deserts. City council votes, school board budgets, zoning fights, high school scores: these are specific, fact-driven stories that one or two outlets cover and nobody else replicates. An AI answer about a national story can draw on a hundred sources. An AI answer about a specific town can draw on a small handful. In a pay-per-value world, that is the only position worth holding.
As to leverage, local publishers do not have it individually, but they have something national publishers mostly lack: trade groups with the means to negotiate on behalf of hundreds of members at once. These trade groups are already negotiating group rates for Google's data and compute products for their members, so moving to group sales may not be such a stretch. For example, the News/Media Alliance has signed opt-in collective licensing frameworks with ProRata and Bria for its 2,200 members, pitched explicitly at local and niche outlets that cannot strike deals alone.
The SPUR coalition may be our collective best path, which started with the Guardian, FT and BBC, added the New York Times and a long list of trade groups this week, including the UK's Independent Community News Network. Denmark went furthest: as the Reuters Institute describes it, the entire Danish news industry negotiates with Big Tech as one body through a collective management organization.
Differentiation over negotiation
Last time I said the publishers most at risk from Google Zero are the ones who treat their audience like oil, a flow to be piped in and metered. The same publishers will be the most disappointed by licensing, for the same reason. They are hoping to swap one pipeline for another. A check from Google that replaces a referral from Google is still Google deciding what you are worth.
Licensing is still one of the best answers to this puzzle, it's just not a fair one on its own. Trade groups notwithstanding, publishers aren't really capable of the organized action that would be required to achieve a rich, global deal, and anyway, the barrier to entry is too low. For a general interest publisher competing on undifferentiated coverage, licensing will be merely a Search Console report that produces a hundred or a thousand dollars here or there, the modern equivalent of the default AdSense accounts that generated nickels and dimes.
The way to survive a market that treats you as a commodity is to stop being one. Cover what no one else covers. Build a direct relationship with the people who read it. And when it comes time to deal with a company much, much, larger than you, do not walk in alone.
