I’m in London for the rest of the week. Looking forward to doing TRB Audience Summit here in February — and spending the runup to the U.S. midterm elections elsewhere.
Today:
Media 2031
Publishers get AI-pilled
AI resistance is futile
Reminder: Join TRB Pro for access to our What’s Working case studies from leading publishing executives. Join now.
PRESENTED BY OMEDA
The most valuable audience you’ll reach this quarter is one you already have — if you can read the signals.
Upsell and cross-sell don't fail because of lack of effort. They fail because the signals are scattered and no one can see them.
See how engagement, content affinity, lifecycle, intent, and more can reveal your next revenue opportunity with Omeda's new guide: Signals to Revenue: Upsell & Cross-Sell Guide for Media. It's the perfect playbook for audience directors and managers in media & publishing.
Included in the Guide:
The Expansion Signal Taxonomy
A Signal-to-Action Framework
The Expansion Signal Audit Worksheet
Media 2031
Axios CEO Jim VandeHei gave a cogent explanation of how Axios is responding to the rapidly changing media environment, with the conviction that publishers need to radically rethink both distribution and creation.
“We envision media consumption shifting to new AI platforms and devices that know us as intimately as we know ourselves,” he wrote. “Our lives and minds will be curated beyond imagination. Our knowledge, habits, tastes and work will fully fuse into our screens, or pins, or bracelets.”
The outline echoes some of the hypotheses laid out by The Atlantic CEO Nick Thompson and The Guardian CEO Anna Bateson in our discussions at TRB Audience Summit, as well as conversations at our recent private salons and roundtable discussions.
Safe bets:
The information market will be K-shaped. All markets are following the current wealth distribution where the top separates from the bottom. Expect more unique information and experiential products geared to the the top part of the K. Axios is talking like a B2B publisher with Jim’s promise to spin up publications for specific job titles. Publishers will scramble to be essential to high-value audience segments.
Focus on unique data sets. News publishers get cranky when journalism is reduced to data, but it is data – and to LLMs, the most valuable data is the data not publicly available. Publishers will increasingly license unique data to AI companies – and many of the publishers of this data will be beyond the media industry.
Go headless. Businesses that depend on introducing friction by controlling an interface with a customer are doomed. Some will have the leverage to muscle through, but most will connect with people through an MCP or connector or API. As one publisher put it: “A year ago we thought we’d build personalization into our own app. Now we realize people will just do that in Claude.” The site and app in this construct become a “lighthouse experience,” like Google’s Pixel phone for Android. Axios Direct is how Axios plans to license feeds of its journalism directly to clients building their own LLMs. This will play out in entertainment as well as information, as hubs like Instagram, TikTok and YouTube become the new TV and media companies jostle with creators as suppliers.
Get human. Individuals will have a leg up by establishing human connection. One participant at our AI dinner last week noted that for all the advances AI has made in coding, it’s still not funny. That also means more events and gatherings. Axios already does over 200 events a year. It hosts an Axios Social Club for its local news outlets. Human stories will always find a market. We are social animals. The Boston Globe story about the western Massachusetts lesbian bar’s battles over masking has frequently been cited as timeless storytelling rooted in humanity. “We will be obsessing about our relationship with human beings,” Anna Bateson told me at TRB Audience Summit.
We’re kicking off a project called Media 2031, which looks at what the media industry will be like in five years. When I ask this question, I typically get an audible sigh and exhale, if not a shrug. But I don’t see how you can operate a media business without hypotheses about how the landscape will alter.
I’ll have more about this project, which will feature podcasts, a survey and an end-of-year gathering in New York City in the coming weeks. Get in touch to discuss partnerships opportunities for this project: [email protected].
PRESENTED BY BEEHIIV
TIME’s newsletters reach millions of readers every week. But until recently, they were stuck with clunky workflows, limited insights, and flat engagement. After migrating 13 newsletters and millions of subscribers to beehiiv (with zero downtime) things changed fast. Open rates on “Inside TIME” jumped 63.8%. Click through rates across their portfolio increased to 11.7%. Editors now publish faster, with better tools and real-time insights. TIME proves that even the most established media brands can evolve and thrive when they own their email strategy.
Publishers get AI-pilled
I’ve noticed a shift in tone and substance of conversations at TRB events and gatherings the past year when it comes to AI. There’s more tangible use cases, more impatience with holdouts and paradoxically more scrutiny on token spending.
Three takeaways from our recent virtual roundtables and breakfast and dinner salons:
More traction. One publisher reported its tech group has tripled its output. But the biggest shift is in non-technical staffers using AI tools to spin up dashboards and prototypes. Another product leader said using Claude to mock up a prototype allowed go from initial idea to launch on a new product in three months. The difference I’ve noticed is that AI is being used across the organization, from newsroom to audience to product to finance to sales (somewhat). “It’s like finding a magic wand,” said one publishing exec. This is in contrast to just the spring, when many were still reporting a focus on basic tasks like summarization, podcast transcripts and SEO tagging in the CMS.
More impatience. One AI lead at our recent dinner put it bluntly, after pleading they didn’t want to “sound like I’m an asshole.” Here’s my paraphrasing of their point: “No one has a right to be a designer. No one has a right to any job they want. You have to earn it, keep practicing it, keep learning and growing. If you want to earn it, you have to use the things that give you leverage, and this thing leverages intellectual power. If we’re farmers and you like working the field with your hands, fine. I've got my tractor and I'm going to outfarm you. I'm going to outdesign you if you don't use it. There's no stopping that.”
More scrutiny. There is a divide in companies between those who get access to the most powerful models and those left to make do with Copilot. One publisher monitors token use and output and will revoke access to the most powerful models if workers aren’t using it effectively. A big publisher built an internal platform to rein in staff vibe coding on servers IT didn’t know about. A wrinkle: big companies struggle with long-term software deals with big tech companies that dictate their use of AI tech rather than using what tool is best for the job. Some also report that legal steps in to make sure the newsroom isn’t using tools like Gemini Notebook for data governance reasons.
Resistance is futile
On the latest episode of People vs Algorithms, we discussed how smaller organizations and individuals have an early advantage in AI adoption as larger enterprises work through governance issues and bureaucratic red tape.
In last week's What's Working: The Atlantic is getting paid by the query by agentic search platforms.
Read the case study to learn:
Why flat licensing fees cap the upside for publishers
How The Atlantic's usage-based deal with Parallel's Index decides what it gets paid
The contract terms The Atlantic uses to protect itself when a data reseller oversteps
Read the case study · Not a member? Join TRB Pro
Send me a note with your feedback by hitting reply.
For sponsorship information, contact Richard Jamieson at [email protected].