Addy Osmani, director of engineering at Google Cloud AI, published new guidance on Agentic Engine Optimization (AEO), a model for content optimization using AI agents.
He put this AEO (not to be confused with an Answer Engine) with SEO, built for systems that download, analyze, and act on content automatically.
This is what he saw. AI agents collapse multi-step browsing into a single request. They don’t scroll, click, or engage with the UI — they get what they need quickly. That makes many traditional engagement metrics ineffective.
A token problem. Osmani highlighted token limitations as the main obstacle shaping content functionality. Large pages can overflow the agent’s context window, causing:
- Discontinued information.
- Skipped pages.
- Hallucinated effects.
Takeaway: token count is now the primary metric for optimization.
Content needs to change. Osmani recommended reorganizing the content of how the agents read:
- Submit answers early (preferably within the first 500 tokens).
- Keep the pages compact and focused.
- Avoid long presentations and buried details. (Agents have “limited patience” for this, he noted.)
Put down on top of the HTML. He also recommended using clean Markdown alongside traditional pages.
- Markdown reduces navigational noise, text, and layout, making content easier and cheaper for agents to analyze.
- This includes making the .md versions directly accessible and available.
Acquisition and structure. Osmani identified emerging patterns to help agents discover and consume content:
- llms.txt as a structured index of documents.
- skill.md files to define skills.
- AGENTS.md as a machine-readable entry point for codebases.
These act as shortcuts for agents deciding what to learn and use.
Why do we care. This adds a new layer of optimization around SEO. If agents can’t parse your content effectively – due to token limitations, layout, or format – they may skip, truncate, or misinterpret it. That directly affects whether your content is used, cited, or processed by AI.
Between the lines. To be clear, the type of AEO Osmani mentioned in his article is not related to Google Search or organic search ranking. Of note, Google’s John Mueller recommended against landing pages and Google does not use the llms.txt file.
- Osmani’s article highlights how AI systems interact with the web and what “optimized” content might look like in that environment.
- AEO shifts the focus from driving visits to enabling successful outcomes within AI workflows.
The subject. Agent Engine Optimization (AEO)
Celebrate deeply:
Search Engine Land is owned by Semrush. We are committed to providing the highest quality of marketing articles. Unless otherwise stated, the content of this page is written by an employee or paid contractor of Semrush Inc.
