I had a call recently with a long-standing publishing client. One of their sites — a project we’d invested heavily in — is shifting commercial direction, which prompted a review of the search strategy. But the conversation quickly moved beyond domain strategy and onto something much bigger: what happens to content businesses when AI systems simply extract and synthesise their work and serve it back without a reason to click?
The economics of publishing, as most sites have known them, are starting to unravel. This is my attempt to think through where that leads.
Change was coming anyway
To be honest, this particular site always felt like a slightly uneasy fit. The editorial team was strong, but the content that performed best was the stuff nobody really wanted to write: high-volume, SEO-friendly, easily commoditised lists, comparisons, guides. The kind of content that didn’t require that level of talent but got clicks.
So this was a timely opportunity to pause and think seriously about the future of content — particularly with the rise of large language models and Google’s AI Mode. These systems now extract and synthesise content to answer user queries directly on the results page. Fewer clicks back to publisher websites. And when the business model is built on traffic — whether for ad impressions, affiliate revenue, or subscriptions — removing the click breaks the entire system.
Without clicks, the economics of publishing start to unravel.
Is the model actually broken?
I keep coming back to this question: if there’s no reward for publishers to write content, why would they write it? The AI systems need this content to keep learning, but they’re effectively biting the hand that feeds them.
This reminds me of the shift from print to digital. Back when the internet became mainstream, newspapers fiercely resisted putting their content online. Why give it away for free? People will stop buying the paper. And they were right — people did. But the mistake was thinking the problem was about selling newspapers. It wasn’t. It was about finding a way to make money online where distribution was free and clicks became the currency.
Publishers adapted. They moved online not because they wanted to but because they had to. Eventually most landed on the same model: ad revenue, affiliate deals, sponsored content for free users, and subscriptions for premium content.
Publish content. Rank in Google. Get traffic. Monetise.
It wasn’t elegant. But it worked for a long time.
What’s the new revenue mechanism?
The move from print to digital felt, in hindsight, almost liberating. Anyone with something to say could publish, grow an audience, and make a living. But this shift — from digital to generative — feels different. Rather than opening things up, it feels like gatekeeping. The big platforms consume content, summarise it, strip it of attribution, and serve it back without a reason to click. They take the value, keep the user in their ecosystem, and publishers are left with the cost.
It’s not sustainable. Publishers need a new model, or at least a different mindset. What that looks like isn’t entirely clear yet. The economics are broken and the next system isn’t fully visible. That’s the uncomfortable truth right now.
Where do we go from here?
Here’s where I’ve landed:
1. Rethink the role of search. Search is unlikely to be the click machine it once was. But it still plays a role as a visibility layer — a brand surface. Rather than optimising purely for traffic, you optimise to show up in the places where your audience first encounters you, even if they don’t click. Think of it more like digital PR. Being cited, even lightly, becomes brand placement. Getting your name mentioned, your product surfaced, your expertise visible — that’s the objective. We measure it by visibility, not just volume.
2. Consider a two-layer content strategy: open versus owned. LLMs are going to keep scraping and synthesising open web content. Rather than fighting it, leverage it.
Layer 1 — open web content: give LLMs your how-tos, your explainers, your foundational content. Think of it as your free sample. It establishes topical authority and signals expertise. It’s the bait.
Layer 2 — owned, moated content: behind that, build real depth. Premium analysis, community-driven insight, proprietary data, strong opinion. Deliver this via email newsletters, paywalled content, member exclusives, or private communities. Give enough away to earn trust — but keep the real value behind your own walls.
I appreciate this creates some tension. The highest quality content that proves your authority is also the content you’re keeping closed. If you hide it, how do you establish yourself as a trustworthy expert? Maybe the answer is uncomfortable. Maybe you need more faith that if you open up your expertise, do right by the reader, and focus on genuinely helping, you’ll be rewarded. Not always with clicks. Not always with attribution. But with recognition over time.
3. Don’t bet it all on search. Sites that rely solely on search are going to struggle with this shift. It’s been brewing for years and it’s accelerating. Any business that doesn’t invest in brand building and diversify its traffic sources is on borrowed time.
Some channels worth prioritising: email (still the most reliable channel you own), video (harder to commoditise and increasingly preferred by younger audiences), podcasts (niche loyalty, multi-task-friendly), community spaces on Reddit, Discord or Slack, and syndication partnerships with non-competitive sites in your vertical.
Revenue model options for publishers
| Revenue Model | Viability | Notes |
|---|---|---|
| Display advertising | Declining fast | AI overviews reduce click-through to pages |
| Affiliate marketing | Under serious threat | AI summarises products well enough to bypass links |
| Paywalls and subscriptions | Viable | Needs strong brand and genuinely differentiated content |
| Sponsored content | Viable | Reputation risk if overdone or misaligned |
| Email list monetisation | Strong | Owned audience is resilient to platform changes |
| Events, courses, webinars | Growing | AI cannot replicate live interaction and community |
| Data licensing and API access | Emerging | Requires legal clarity and industry cohesion |
| Donations and crowdfunding | Niche | Sustainable only for passionate, loyal followings |
Where I’ve landed
There’s an inflection point coming, and it’s going to be both fascinating and uncomfortable to watch. Either the platforms that scrape, summarise, and repackage content will find ways to return value to the people creating it, or we’ll see more content go private — paywalled, gated, or withdrawn from the open web entirely.
My instinct is that the latter is more likely. The web will fracture into two layers: one of surface-level, commoditised information that AI systems surface for free, and another where the real thinking, nuance, and value lives behind closed doors.
Until there’s a viable revenue model for open content in the generative era, visibility alone won’t be enough. Publishers will need to build ecosystems that don’t depend on Google to stay alive. Which is, ultimately, probably a good thing.