Grandma’s always right.

This guy’s 78-year-old grandmother just explained the new world of content discovery in one brutal sentence: "Why visit websites? Google told me everything."

That doesn’t even refer to search links, it refers to AI overviews. For the last 6 months, we’ve been writing obituaries for content publishers. The eulogy was simple: educational content = AI will summarize. Entertainment content = hanging in there but reduced. Only shopping content survives, for now…

As an ex publisher, broadcaster and pub tech stalwart, I have a soft spot for seeing publishers survive and thrive with strong adaptation.

And yesterday was the start of a new paradigm in publisher asset control.

Cloudflare set off a reversal that Chris Long calls "earth-shattering" - may be a slight oversell if you’re not a publisher but it is big. Every new domain on Cloudflare now blocks AI crawlers by default. But here's where it gets spicy: publishers can now charge micropayments for every. single. crawl.

Let that marinate.

We know that OpenAI's crawler scraped websites 1,700 times for every one AI answer. Anthropic? 73,000 times. That's clearly not a partnership - that's digital colonialism.

Beyond Pay-Per-Crawl

Anthony Katsur from IAB Tech Lab sees it clearly: pay-per-crawl is just the appetizer. Pay-per-query is the main course. But there’s even more.

This isn't about defending against zero-click searches. It's about building something AI can't live without: your proprietary data empire.

The average enterprise averages between 172-664 internal applications, which scatters data like digital confetti. Publishers sit on goldmines of contextual insights & audience intelligence. It’s content that AI desperately needs but can't create.

The Hidden Opportunity

We're witnessing the birth of “content-as-a-service” meets “data-as-infrastructure”.

Publisher articles aren't just pageviews anymore. They're training data with price tags attached. Audience insights aren't just ad targeting, they're the proprietary intelligence AI companies will pay a premium for.

Smart publishers will:

• Transform content archives into specialized AI training sets

• Create a stable, durable and fully owned ID without needing your email address

  • And use this to build first-party data cooperatives for collective bargaining

  • Create Conversion APIs that goe toe-to-toe with Meta and aren’t sellotaped together

  • Immediately increase CDP coverage to 100% of visitors

• Create synthetic datasets from their unique knowledge bases

• Develop API-first content strategies where every query has a price

As Rand Fishkin's prisoner's dilemma warns: someone will always give content away free. But commodity content was always worthless. Content people stay loyal to creates proprietary data, and that's irreplaceable (unless it’s not turned in to a data fortress).

Publishers need to fortify

20% of the web runs through Cloudflare. And I can guarantee that it’s the good 20%.

Publishers who use this news to move by building their data fortresses now to experiment with these new models - they'll own the future.

We spent six months believing publishers were dead. Now it feels like time for the publisher empire to strike back.