A new best practice for "Agent Experience" is using content negotiation to serve different payloads to AI agents. When an AI crawler requests a page, the server can respond with raw Markdown instead of rendered HTML, significantly reducing token consumption and making the site more "agent-friendly."
Related Insights
As AI agents become primary consumers of documentation, the battle for superior developer experience shifts from visual design to content accuracy. An agent reading raw markdown doesn't care about UI, making the underlying information paramount and the foundation of a modern DevEx strategy.
The rise of AI browsers introduces 'agents' that automate tasks like research and form submissions. To capture leads from these agents, websites must feature simple, easily parsable forms and navigation, creating a new dimension of user experience focused on machine readability.
Websites now have a dual purpose. A significant portion of your content must be created specifically for AI agents—niche, granular, and structured for LLM consumption to improve AEO. The human-facing part must then evolve to offer deeper, more interactive experiences, as visitors will arrive with their basic research already completed by AI.
Instead of guessing how to make your site more compatible with new AI browsers, directly ask the AI itself. Prompt ChatGPT with your URL and ask what changes are needed on your site to ensure the right answers appear when users search with the Atlas browser.
The audience for marketing content is expanding to include AI agents. Websites, for example, will need to be optimized not just for human users but also for AI crawlers that surface information in answer engines. This requires a fundamental shift in how marketers think about content structure and metadata.
To make product and service pages AEO-friendly, marketers should add specific structural elements. Including a 'TLDR' section, an accordion-style FAQ based on buyer questions, and direct competitor comparison content helps LLMs easily parse and surface key information.
The first step to influencing AI is ensuring your website is technically sound for LLMs to crawl and index. This revives the importance of technical audits, log file analysis, and tools like Screaming Frog to identify and remove barriers preventing AI crawlers from accessing your content.
When building multi-agent systems, tailor the output format to the recipient. While Markdown is best for human readability, agents communicating with each other should use JSON. LLMs can parse structured JSON data more reliably and efficiently, reducing errors in complex, automated workflows.
In the era of zero-click AI answers, the goal shifts from maximizing time-on-page to providing the shortest path to a solution. Content must lead with a direct, data-dense summary for AI agents to easily scrape and cite.
The rise of AI agents means website traffic will increasingly be non-human. B2B marketers must rethink their playbooks to optimize for how AI models interpret and surface their content, a practice emerging as "AI Engine Optimization" (AEO), as agents become the primary researchers.
