Image Gareth Glaser https://www.linkedin.com/pulse/appeal-open-internet-gareth-glaser-kwnbe/
The open web – free content written by actual humans about things actual humans care about – has been in decline for more than a decade. I’ve had a front row seat throughout – first at Federated Media, which was built 20 years ago to support independent publishers, then on the Board of Sovrn, which continued Federated’s work on the programmatic/data side of the publishing business. I’ve also taught and practiced journal…
Image Gareth Glaser https://www.linkedin.com/pulse/appeal-open-internet-gareth-glaser-kwnbe/
The open web – free content written by actual humans about things actual humans care about – has been in decline for more than a decade. I’ve had a front row seat throughout – first at Federated Media, which was built 20 years ago to support independent publishers, then on the Board of Sovrn, which continued Federated’s work on the programmatic/data side of the publishing business. I’ve also taught and practiced journalism for the past few decades, and started and advised countless ventures that depend on traditional media revenue streams.
In short, I know it ain’t pretty out there for advertising-supported publishing. Social media dug the grave, and now the nail gun of generative AI feels seems to be merrily fastening the lid over the open web’s pine box coffin.
So is there anything to be done? I’ve noticed two distinct approaches to the problem, but only one of them strikes me as having any chance of helping.
The first is what I’ll the the “biz dev solution.” This is where the big platforms and AI companies strike deals with large media companies, bringing revenue to the large media companies, and fresh content to the AI/platform players. We’ve seen these deals for years now, but they’ve picked up considerably in the age of consumer AI. They’re a bit of a devil’s bargain – never enough money to truly change anything for the media companies. For the AI companies, who have a voracious maw to feed, they’re a necessary evil. These kinds of deals take forever to negotiate, and they don’t scale – sure, CNN, Disney, and Reuters have the resources to cut these deals, but the mid and long tail of the Internet, who make vastly more content and have far fewer resources? They get nothing.
The second approach is to build revenue sharing for publishers directly into the Internet’s core infrastructure – which for simplicity we’ll call the “content marketplace” solution. This approach has been championed by companies like Cloudflare (see my write up of their approach here), and Microsoft is noodling on the idea as well (as are many others). Seen from space, the idea behind an infrastructure approach is to leverage neutral protocols to enable every player in the content ecosystem to set and negotiate pricing in real time.
If that sounds a lot like the world of programmatic advertising, you’re not wrong. Many in that industry are excited by the potential to adapt their real-time bidding and inventory management infrastructure to the problem of accurately and fairly valuing content in an age of AI – you can read a good overview of the potential for such a system in this recent post from Ari Paparo.
As you can imagine, I’m a much bigger fan of the content marketplace approach – it feels much more scaleable, and potentially far more valuable to smaller players who lack the resources to directly negotiate with companies like Google or OpenAI. Biz dev driven approaches create a fragile, consolidated ecosystem dominated by large companies and their established business models and incentives. A true content marketplace would allow an unlimited number of parties to add and extract value, as well as lay groundwork for unexpected innovation across the network. In short, it’s a far more robust and sustainable system.
But will there be enough demand – and revenue – for content marketplaces to actually become liquid and self-sustaining? Programmatic advertising became a hundred-billion-dollar market because there was massive demand from advertisers. Will there be similar demand for quality content to feed AI models? Unlikely, unless the AI industry proves it can drive far more revenue through its ecosystem than it currently does. And the only way that will happen is through advertising.
Which brings up an interesting, if rather obvious observation: The locus of advertising engagement is shifting – from publishers and content, where it’s lived since the birth of modern advertising – to AI-engagement surfaces like chatbots and other AI-driven apps. Search was the first step in that shift – Google was not a traditional publisher, per se, but it leveraged publisher content to become the fastest growing and most successful advertising business in history. Social media ran the same playbook – dis-intermediating publishers large and small. Now Google (and Microsoft, along with some smaller players) are showing ads in AI-driven search and chat results. Meta, OpenAI and others will soon follow.
Will they share that revenue with publishers? Certainly not without being forced to. Which is why I am cheering the Cloudflares of the world, who are giving publishers the tools they need to deny AI engines the fuel they need to win. I’ll be watching this space closely over the next year. The future state of the web depends on how it all plays out.
—