About This Episode

I sat down with Sean Barber on Search With Sean for a long conversation about AI agents, websites built for machines, and how to cut through the hype that surrounds all of it. We started with the practical question of what an AI agent actually does when it lands on your site and what it needs to succeed there. From there we got into the uncomfortable fact that most AI systems don't render JavaScript, and what that breaks for sites that depend on it.

We also talked about whether GEO is really a different discipline from SEO, and why solid foundations like clean HTML, schema, and semantic structure still decide who wins.

The second half went beyond tactics. We covered the measurement problem when clicks and "sticky" metrics stop making sense, why no single team owns AI optimisation yet, and why informational sites are being hit the hardest. I explained how I test agents on real sites using server logs and agentic form testing, and why agents get stuck in places where humans don't. We closed with predictions about the web splitting into open and gated, why LLMs only ever recommend three or four brands, and why you should run from anyone promising to rank you number one in ChatGPT.

Key Topics Discussed

• What an AI agent really does on a website, and what it needs to succeed
• Why most AI systems don't render JavaScript (and what that breaks)
• SEO vs GEO: whether it's actually a different discipline
• Misinformation, YMYL and hallucination, and how much of it you can control
• Writing down your canonical brand identity and giving it version control
• Whether AI is genuinely adding value
• Measuring success when clicks and "sticky" metrics stop making sense
• Why no team (dev, SEO, UX, CRO) owns AI optimisation yet
• Why informational sites are hit hardest, and giving people a reason to visit
• Tracking AI agents with server logs and spotting silent failures
• Agentic form testing: why agents get stuck where humans don't
• The web splitting into open and gated (pay-to-crawl, trust and authenticity)
• Why LLMs only recommend three or four brands, and what that does to competition
• Lessons from 200+ episodes of No Hacks