Capitulation

Negotiating terms of surrender

Reminder: Sign up for updates about The Rebooting’s full week in Cannes if you’re one of the fedora’d people who will be hustling up and down the Croisette. 

Today's admittedly long newsletter:

  • A new PvA about AI hallucinations
  • Highlights from yesterday’s audience development online forum
  • AI in the content creation process
  • For TRB members: The capitulation phase of publisher-tech relations

AI hallucinations

On this week’s episode of People vs Algorithms, we discuss AI hallucinations from the standpoint of the hallucinations humans have about AI’s impact. Some hallucinations discussed:

  • AI ending web publishing and advertising
  • Publisher AI deals having a different ending this time around
  • Google rolling back AI Overviews because of some viral slop
  • Vivek Ramaswamy turning BuzzFeed activist investor

Listen to PvA on Apple | Spotify | other podcast platforms


Where’s the traffic?

Building and developing an audience in the fragmented media landscape of 2024 requires interlocking strategies to efficiently find high-value audience segments. The current tumult in the SERP brings home how much audience development has changed, coming on top of social traffic declining. 

Yesterday, in partnership with AI-powered social media management platform True Anthem, The Rebooting held an online forum focused on the new paths to building audience at a time of distribution tumult. I was joined by True Anthem CEO Chris Hart, a360media vp of audience Lorraine Goldberg and Newsweek svp of audience Josh Awtry. Some takeaways:

Traffic isn’t dead. It’s easy to say the traffic era is over, but traffic is still the lifeblood of media business models. Traffic is simply distribution. Those patterns are changing, but publishers cannot opt out. The challenge is finding new sources of traffic that can be converted into audience.

X is mostly dead. Even for Newsweek, X is on the wane. For a360media, home to entertainment brands like US Weekly, X is not a major factor. Twitter historically was never a critical traffic source for most publishers, and that has only accelerated in its current guise.

There’s no one algorithm. Algorithmic distribution is here to stay. The follower model is over. That means publishers will be at the mercy of increasingly sophisticated algorithms that are tuned for individuals. That doesn’t mean social distribution is a matter of throwing content against a wall and seeing what sticks. But it does mean using data signals to double down on what is sticking and amplifying it through recirculation.

Google Discover, Apple News and Reddit remain robust. For current news content, Google Discover is more critical than the SERP, which dominates distribution for evergreen content like recipes and gift guides. Similar patterns exist with Apple News and Reddit, particularly as Google has increased distribution there. Reddit is improving its tools for publishers, but the unique nature of its discussion forums make influencing distribution there indirect.

Resource constraints are the new normal. The temptation to turn social posting over to AI is real but not a viable strategy, according to participants. Instead, publishers need to use AI to better inform the selection of content to promote through different channels and automating as much as possible of the grunt work that was formerly done by social media editors. 

You can watch a replay of the hourlong session on demand. Thanks to True Anthem for supporting the discussion. Check out how it uses AI to help publishers grow traffic and automate social media workflow.


Publishing in an AI era

If you’ll be at Cannes, you’re invited to come to a live podcast I’m doing on Thursday, June 20, with Dotdash Meredith CEO Neil Vogel and Axios senior media reporter Sara Fischer to discuss the state of publishing in an AI era. I’m particularly interested in whether Neil will continue his stance that DDM will never use AI to produce content. Some of this is technicalities. It seems inevitable to me that AI will play a role in the content creation process; it just depends on how it is used and, critically, the quality of the output. The lines will inevitably get fuzzy within all publishing companies. 

Sign up for the event here.


Capitulation

On April 9, 1865, Robert Lee showed up to Appomattox Court House for what was technically a negotiation to end the fighting that his army of Northern Virginia had waged for four years. The negotiation lasted a few hours because Lee was really there to be told the terms of surrender. Lee had no leverage: army surrounded and decimated, supply chains cut, escape routes blocked. And yet, he got a good deal as Grant and Lincoln wanted the war over quickly. 

That’s the hopeful scenario for publishers. They have lined up at their own Appomattox to sign their terms with OpenAI. The trickle of deals by AP and Axel Springer have become a flood as in the past week News Corp, Vox Media and The Atlantic have signed on to “partnerships” that absolve OpenAI of vacuuming their content without consent to build a competitive product. OpenAi now has deals with 70 publishers. Score another win for the outlaws

Jessica Lessin, founder of The Information, wrote in The Atlantic last week about why these deals are inopportune and a long-term error for media companies. There was a time, late last year, when it appeared publishers would dig in for a fight. I sensed the depredations and humiliations of a generation of executives getting screws over by big tech companies was fueling a fighting spirit. Diller talked of a legal coalition that would demand “billions, not millions.” The coalition would include The New York Times, News Corp and Axel Springer. Only the Times remains, and Diller-owned Dotdash Meredith cut its own deal earlier this month.

That was noteworthy because if the exposure the company has with its reliance on evergreen content in many of its properties. The New York Times will carry on with its lawsuit since it has unique leverage afforded by a sustainable business model. What leverage could Vox bring to an OpenAI negotiation? The LLM won’t have new Eater reviews? The realist view is: take the cash.