Digiday Publishing Summit September 2026 Recap: How publishers are rebuilding for the post-search era

Executives also remained cautious about the emerging AI economy: some are holding out on licensing until the economics make sense, while others are building GEO businesses, even though questions around compensation, attribution and standards around AI visibility remain unresolved.

Each Digiday event kicks off with the edit team breaking down current industry themes. Here’s where we think we are:

1. The post-search era is taking shape. Publishers are working to reduce their dependence on Google as AI-powered, zero-click search erodes referral traffic and makes search-driven scale less reliable.
2. Diversification is no longer optional. As traffic-dependent economics weaken, publishers are building more resilient businesses around subscriptions, advertising, video, licensing and emerging AI revenue opportunities like GEO products.
3. AI licensing remains a revenue opportunity for publishers, though the landscape remains inconsistent. Publishers are moving beyond experimentation with AI companies, weighing licensing deals based on revenue potential and control over their content – though larger companies still hold much of the negotiating power compared to mid-size and smaller publishers.
4. The AI shift is reshaping publishing organizations. Publishers are restructuring audience and newsroom teams, creating new AI- and discovery-focused roles, retiring legacy functions and upgrading technology to operate more quickly in a changing distribution environment.
5. Direct audience relationships are becoming more valuable. Publishers are prioritizing brand, loyalty, registration and subscriptions as they balance advertising and consumer revenue and become less reliant on platforms to deliver audiences.

Here are some of the sessions — and takeaways — we don’t think you should miss:

Reuters’ paywall pays off

Reuters’ dynamic paywall is helping the news organization grow subscriptions without sacrificing ad revenue, using machine learning to determine when individual readers are most likely to subscribe versus when they’re better monetized through advertising. The system weighs user behavior, content performance, conversion likelihood and potential ad revenue, and adjusts with the news cycle. Phil Andraos, Reuters’ gm for digital, said the ad business is now performing better than before the paywall, in part because subscribers are highly engaged, authenticated users who visit more frequently and generate additional ad impressions and first-party data. Reuters is also prioritizing monetizable traffic over sheer volume elsewhere: since adopting a block-by-default approach to bots in May, bot traffic has declined while its “monetizable traffic” has remained unchanged.

• “It’s basically like magic. Overnight, the dollars start to flow.”
• “So often we focus on monthly uniques or monthly number of page views, but that I think is the wrong metric. It’s really about engagement. A smaller pool of highly engaged users can drive so much more ad revenue than a high base of loosely engaged readers.”

Future’s GEO business gains ground

Future is adapting its branded content business to the rise of AI through Future Optic, its roughly year-old GEO product, which now has more than 30 clients and is seeing renewals. Future is leaning on its publishers’ editorial authority, consumer expertise and existing brand relationships to sell GEO as an extension of branded content. CRO Mike Peralta said the company is currently measuring campaigns primarily through brand visibility and publisher citations in AI answers. But he acknowledged that the industry has yet to solve the bigger attribution question: whether greater AI visibility actually translates into purchases, sales or improvements in brand metrics. Peralta argued that changing discovery habits could ultimately create an opportunity for publishers, even while audience-dependent revenue streams like programmatic advertising and e-commerce face pressure. 

• “More is going to be required of us as publishers… the more that we’re able to come to the table with a certain level of expertise, not just within our own content, but across the whole ecosystem, I think it’s going to be better for the client.” 
• “LLMs [are] still largely black box, and so we have to figure out, how do we connect those dots? So how do we write for our consumers, but it has to be understood by the machines.” 

How The Wall Street Journal’s push to move beyond text is reshaping newsrooms

The Wall Street Journal recently made two significant changes to its podcast operations with another shift in the works. It hired Samantha Henig to a newly created head of multimedia role, a position overseeing the Journal’s video team as well as its audio team. But leading the latter team is not a head of audio. That had originally been the plan, but the publisher ended up re-titling the position to senior director of shows, according to Taneth Evans, head of digital at The Wall Street Journal. The position, held by Leital Molad, is primarily concerned with audio, but not exclusively. To that end, the Journal is also adopting a show-centric strategy with its podcast operations.

• “[The head of multimedia position is] a new role, and it’s kind of us acknowledging that we need a conductor atop all of these formats because they need to be able to intermingle and you know be nimble with each other.”
• “There will certainly be audio producers and all of the traditional teams that go into an audio production [that are part of the new show-centric satellite teams]. There will probably be video producers. Maybe there will be people who are dedicated to the social output of that show or the newsletter output. There might be people from live events tapping in if we’re putting it on stage.”

Inside Politico’s AI newsroom overhaul

Politico is bringing its newsroom, product and AI strategies closer together while rebuilding its underlying technology to support AI-driven workflows and products. Francesca Barber, evp of product, audience and AI, described her dual role across the newsroom and business, and how Politico is experimenting with AI tools. One of Politico’s biggest bets is replacing its traditional CMS with Sanity, a content operating system that treats journalism as structured data, allowing the system to proactively surface relevant reporting, sources and information to journalists and eventually distribute Politico’s reporting more easily across products and platforms. Politico is also working to create a more unified app and user experience across the U.S. and Europe. 

• “I do believe that this is a moment with AI where it’s creating content — it’s so close to the writing, the discovery of information — that I felt it was very important that I had a sort of dual role reporting into the global editor in chief and the CEO.”
• “We’re creating an authoring tool that is proactive. It can proactively alert you. It can proactively draft something for you relevant to one particular area of your story.” 

How The New York Times evaluates AI licensing partnerships

The New York Times is taking a selective approach to AI licensing, evaluating potential deals around three principles: a fair value exchange, a sustainable partnership that goes beyond a one-time payment, and control over how and where its content is used. But Adam Greenberg, the Times’ vp of strategic partnerships, said the emerging AI licensing marketplaces remain “underdeveloped,” with unresolved questions around content valuation, usage and whether the revenue is meaningful enough for publishers. However, Greenberg said partnerships can also give publishers more leverage to push AI companies for better attribution and referral mechanisms. Greenberg said relatively few AI deals have been signed because those conditions remain difficult to meet.

• “I think where we end up is very unclear… It’s very hard to get these things going. It takes years, and so a marketplace that is valuing content in real time for use in AI training or inference is just going to take a long time.”
• “The reality is that there are not that many deals that have been signed in the last few years… because those conditions are tough to meet in a lot of different ways… The devil is really in the details.”

USA Today Co.’s publisher growth playbook: AI, audiences and revenue

USA Today Media has been among the most vocal publishers on the potential to block Google’s crawlers. And for good reason. Google’s share of referral traffic to the news organization’s sites has dropped from 70%-plus a few years ago to roughly 40% now, according to USA Today Media president Kristin Roberts. That’s still a sizable amount, but the drop helps to explain why the company is prioritizing alternative audience and revenue sources, such as referral traffic from Instagram’s Threads, TikTok and Reddit as well as content licensing deals with Meta and Microsoft.

• “I mean, I would love to [block Google’s crawlers] immediately, right? We’re not going to do it immediately, but we are at a place now where that traffic is not coming back. And they are one of the only crawlers we don’t block, and we don’t block them because we still rely on search. But to a decreasing extent. And so as soon as we don’t need to rely on our articles being indexed, then we should block them.”
• “These AI deals are lumpy, and that’s because [USA Today Media executives are] still trying to figure out where are the lines and how do you value things. Each deal has been valued differently, but we consider each deal almost a floor. What I liked so much about the Microsoft deal was that it was a marketplace. You could see is my content performing in this space, and if it wasn’t, how do we adjust for that.”

CNBC bets on AI — but not licensing

CNBC has not signed any LLM content-licensing deals and has maintained a “hard line” that AI companies must adequately compensate it without cannibalizing its business, said Sally Shin, CNBC evp of growth and partnerships. But CNBC isn’t anti-AI. The publisher is using data partnerships, AI and personalization to build a stronger direct relationship with audiences. Shin said the publisher is looking for partnerships that give its professional and investing audience an “edge,” pointing to its multi-year Kalshi partnership, which incorporates prediction-market data into CNBC’s journalism. At the same time, CNBC is exploring AI internally to make its large archive of interviews and video more useful.

• “The reason why we haven’t done any deals with any of these LLM companies is because we want to be compensated for the content that we’re producing, and it’s not cannibalizing our business.”
• “Talent and compute labs are spending a lot of money, millions of dollars per employee, billions of dollars in terms of compute. So where we stand is that the labs should also be paying the same amount of money for the data that they are ingesting into their LLMs.” 

How The Verge is building audience for a zero-click future

The Verge isn’t just preparing for Google Zero — it’s already operating as though Google traffic is incremental. Instead of optimizing primarily for SEO, virality and pageviews, the audience team now acts as the “keeper of the funnel,” focused on acquiring high-value readers and turning them into registered users, newsletter readers and subscribers, said Esther Cohen, director of audience and subscriptions. The Verge is shifting its audience strategy away from maximizing pageviews and toward building a smaller but more valuable direct audience. Central to that strategy is a stronger homepage and a reorganized, larger audience team.

• “I think of Google traffic now as like extra. I don’t even like to plan for Google or optimize for Google.” 
• “If our overall revenue, our overall traffic, our overall engagement, subscription, registered users number is growing healthily… and Google’s declining as a share of our revenue as much as our actual traffic is declining from Google, I’m OK with that.” 

From Axios’ Smart Brevity to AI visibility

Axios reached its revenue goal for 2026. Now the news publisher is building a new revenue stream to keep up the pace in 2027 and beyond. Axios is taking to market Axios Direct, a new product that will provide Axios’s reporting in the form of content feeds that effectively take the concepts of a Bloomberg terminal and an RSS feed and update them for the agentic AI era. Axios Direct will eventually take the form of three separate feeds aimed at three types of customers, according to Axios chief revenue officer Jacquelyn Cameron.

• “We believe that there are three pathways where we could make net-new revenue impact to the bottom line. The first pathway is to working with clients who will take an action off of that information. We think of them as investment companies with high assets under management.”
• “Year over year, we are posting a significant double-digit growth yet again, and much earlier this year than what I had originally anticipated. Last year we hit our number at the end of October, and this year we hit it early September.”

Ziff Davis bets on trust over traffic

Ziff Davis’ Steve Horowitz, president of the tech & shopping division, stressed that publishers shouldn’t try to replace lost Google traffic. Instead, its strategy is to diversify the “legs of the stool” across platforms and revenue streams, including direct audiences, social, premium advertising, affiliate commerce, licensing and monetizing proprietary data. The company is diversifying both where its audiences come from and how it makes money from them. (For example, Ziff Davis audited its archives, including information that had never appeared on the web, to explore how that data could support new products or even be sold.) Horowitz said roughly 35% of Ziff Davis’ revenue is associated with web traffic and about 17.5% with search, making the impact meaningful but not “catastrophic” across its broader portfolio. Meanwhile, Ziff Davis is generating revenue from clients looking to improve their AI visibility, although success metrics vary widely by advertiser.

• “We all kind of thought SEO was owned. We felt that we kind of had a right to that… Well, that’s kind of been demoted as a business.” 
• “I still think we all need to figure out attribution… especially in the world of AI, when we’re all being shown up here and we are influencing down here… That attribution loop is not clear right now.” 

How Semafor is redefining the head of video role

Semafor technically had a head of video before Adam Banicki stepped into the role this summer. But the role has changed with Banicki’s appointment. Instead of being strictly a newsroom role, Banicki’s position spans the editorial and business sides of Semafor. As Semafor’s head of video, Banicki reports directly to co-founder and editor-in-chief Ben Smith but “with a dotted line to Rachel Oppenheim, our chief commercial officer,” he said.

• “My first three months, I think, was really getting involved in the commercial side, helping shaping go-to-market decks, thinking about pricing strategy, joining kickoff calls to help close deals that were tied to video.”
• “With Ben, you catch him where he is. I sit right next to him. I will just swing by his desk, or he’ll swing by mine pretty routinely, and then he has a more formal meeting with all of his leadership team. And then with Rachel, I have a standing, and I’m in all of her leadership meetings.”

Puck’s bet on creators as the growth engine in AI era

Puck is attempting to redefine the relationship between journalists and the business side. Instead of primarily measuring writers against traffic, it gives them equity and ties compensation to subscriber acquisition and retention, then builds additional revenue products around their individual strengths. Puck co-founder Liz Goff outlined the model: journalists can earn additional compensation based on subscribers they acquire and retain, as well as their participation in products such as events and podcasts. That model extends to product development, too. Editorial, marketing, technology and commercial teams meet regularly with individual journalists to evaluate subscription performance and develop new products around each writer’s strengths, from ticketed events to podcasts and video. 

• “The whole business was built around the idea that… journalists are at the center of the revenue model, and the journalists and the business people really are partners in building a sustainable model for journalism.”
• “Having the journalists really close to the business helps them do better work and helps us build a better business.” 

Couldn’t be there? Here’s our video on how it worked.

Content licensing deals weren’t the only AI-related accords highlighted during the Digiday Publishing Summit. 

In a live interactive workshop, attendees joined into teams to try their hands at negotiating advertising deals via AI agents. Teams played the part of publishers and were presented with an advertiser campaign brief. Then they were tasked with using AI agents to create the most attractive inventory package in order to secure the highest share of the advertiser’s $400,000 budget and at the highest premium CPMs.

Teams drafted initial prompts to instruct AI seller agents to create inventory packages to submit to an AI buyer agent. The AI buyer agent then countered each team’s offer, to which teams had to draft prompts for the AI seller agents to respond. The AI buyer agent then reviewed the revised offers and dedicated how the advertiser’s budget should be allocated.

Congratulations to Team A for winning the game in a come-from-behind victory (and condolences to Team E, which was the only team to enlist an AI chatbot for assistance and had the lead at the game’s midway point).

Sourced from town halls, sessions and hallway conversations at DPMS:

• “Everyone is coming with a different [AI licensing] deal, and none of us are allowed to talk about these deals publicly… It’s super freaking frustrating.”
• “The money is largely concentrated in the very biggest players. They don’t want to participate through [an AI licensing] marketplace. And we have no ability to see through, nor even talk amongst ourselves, about what ‘good’ might look like. “
• “We can’t even talk about the deals that we sign.”
• “With the Metas and the Googles, you kind of know what’s the standard contract. We all know what’s the rev share. We might not be able to disclose it, but there’s a standard contract from those partners. I think what we’re facing right now – and maybe it’s because of the newness of the industry – is that there is a distinctive contract probably for everyone, and it depends on your ability to negotiate. And I think that is scary because we can’t put a concentrated effort. We just have to survive on our own.”
• “This is a new fresh hell.”
• “Programmatic is going to go down consistent with what we see in our page view decrease. The question is: What is the value that our content has, and what opportunities are available to us to achieve a fair value for that content?… It’s really, really hard to form a point of view on that because of the information control that is on this.”
• “Some of the contract terms can be really, really ugly.”
• “The idea of Google Zero is, we’re not going to hit zero… I think what people mean is search zero. it will decline. It’ll continue to decline. We’ve got to figure out what that means. That probably means we have to monetize better, more efficiently.”
• “I think the days of making money on mass traffic from Google referrals are generally down and over.”
• “The days of… high yield programmatic revenue is just coming to an end.”
• “We’ve really focused on optimizing articles for AI, and we’ve seen a 40% increase in customer acquisition and traffic as a result… It’s not like going from 100 million users a month to 140, but it’s substantial.”
• “Frankly, a personalized newsletter, everybody’s kind of doing it.”
• “If your paywalls are set up correctly, you are largely protected from having those results be surfaced in information that any gen AI is going to be able to leverage.”
• “We continually do research on: Is there a benefit to putting this behind a paywall? What could we gain from that in terms of revenue procession as we see monthly active users kind of diminish in the way that the entire open market is seeing? And the math never works.”
• “People are not landing on our sites because they’re getting the information they need from a Google search response or jumping into an LLM themselves. You’re just throwing gas on that fire by putting any sort of blockage between the consumer getting the information that they’re coming to you for.”
• “We use the registration wall really to get more information on known users, and then we have advertisers who pay to sponsor our newsletters, and so we have really good information about the people reading our newsletters because of this registration wall, which allows us to obviously charge more for that.”
  

1. Search traffic isn’t coming back. Publishers are moving beyond the era of mass, low-value traffic monetized through programmatic ads and accepting the new reality.
2. Publishers are reducing their reliance on search. They are restructuring audience teams and investing in platform partnerships, AI discovery tactics and new distribution strategies.
3. AI licensing is growing, but remains immature. Deals offer publishers revenue and control over their content, but opaque terms and limited transparency remain challenging — and the revenue from those deals isn’t close to making up for the revenue lost from declining referral traffic.
4. Diversification is essential. Publishers are leaning into subscriptions, sponsorships, premium advertising, newsletters and direct audiences, though these businesses don’t always offset traffic-driven losses.
5. AI optimization remains experimental. Publishers are testing structured content, markdown, bot-focused strategies and onsite AI chatbots, but their long-term value remains unclear.

How Reuters’ dynamic paywall is boosting subscriptions and ads

Reuters’ ad business is now performing better than before it launched the paywall two years ago, in part because subscribers are highly engaged, authenticated users who visit more frequently and generate additional ad impressions and first-party data.

Media Briefing: Overheard at the Digiday Publishing Summit, Sept. ‘26 ‘Google Zero’ edition

During the closed-door town hall sessions at the Digiday Publishing Summit, publishers seemed to finally accept their new reality: Google search traffic is not coming back. 

Axios preps new Axios Direct feeds for AI models, agents as revenue tops 2026 goal

Axios reached its revenue goal for 2026. Now the news publisher is building a new revenue stream to keep up the pace in 2027 and beyond, launched a new product that will provide Axios’s reporting in the form of content feeds for the agentic AI era.

How Semafor is redefining the head of video role

Semafor technically had a head of video before Adam Banicki stepped into the role this summer, but now instead of being strictly a newsroom role, Banicki’s position spans the editorial and business sides of Semafor.

What The New York Times looks for in AI licensing partnerships

The New York Times has three core principles it sticks to when evaluating those potential deals, but it’s proven “tricky” to figure out those conditions in AI licensing marketplaces.

Media Briefing: Publishers question whether personalization actually pays off

Publishers are betting that personalized newsletters and sites can squeeze more value out of the readers they already have, but there’s little consensus on whether personalization is a meaningful differentiator, or just table stakes that don’t move the needle on revenue.

Read all Digiday Publishing Summit Summit coverage.