From Google to AI: Your strategy with SEO has shifted

Over the last 2 decades, SEO has revolved around one question: “How do I rank higher on Google?”

Before Google was popular, I was briefly the No. #1 search for Google ranking because I’d written a page “Rules to Better Google Ranking”.

After years of Google dominating information search, new AI tools like ChatGPT, Copilot, and Claude are increasingly becoming the go-to sources for finding information. This shift signals a serious threat to Google’s long-standing dominance in the search space.

Since people are increasingly asking language models for answers instead of searching the web themselves, that means your content isn’t just competing for clicks on Google, it’s competing to even be seen by an AI.

This shift has sparked a new term, “LLM Optimization” (LLMO), a new set of best practices that help AI models discover, understand, and cite your content.

Google vs AI Search growth

In 2024, AI tools account for approximately 6% of the search market, which is small. However, with a projected annual growth rate of up to 35% for 2025, this is important. It is expected that AI’s share of the market will reach 14% by 2028.

At the same time, the data shows that Google searches are not dropping. So, despite AI’s increasing presence, the total search market is expected to grow by around 40% between 2022 and 2028. This suggests that AI is not cannibalizing traditional search but rather contributing to the market’s overall expansion.

Future Proofing your website

While the integration of AI in search is so far not a threat to traditional search engines, websites that are focused on informational content may experience traffic reductions as AI tools provide direct answers to users. Conversely, sites emphasizing transactional content are likely to see continued growth.

So, how do you prepare for the future and get ahead of the curve?

✅ #1 – Use Markdown – Clean, Structured, and LLM-Friendly

Markdown is a lightweight markup language that uses plain text formatting. It’s:

• Easy to read and parse
• Consistent across platforms
• Ideal for documentation sites, blogs, and frontend frameworks, like Next.js or Astro. Works great with markdown-based content management systems like TinaCMS

💡 Why it matters:
LLMs learn better from clean, semantically structured content. Markdown naturally provides this structure, with clear headings, lists, and link formatting, making it easier for models to extract useful context and answer user queries more accurately.

✅ #2 – Add an llms.txt to your website

Inspired by robots.txt, llms.txt file gives you a way to signal which parts of your website are meant for LLMs to crawl and index. You can:

• Point to your most helpful, evergreen content
• Exclude pages that shouldn’t be used in AI training or inference
• Optionally, highlight licensing terms for your content

💡 Why it matters:

This file will become a standard way to opt in or out of LLMs accessing your content, giving you control over how your site shows up in AI-generated answers, just like robots.txt does for Google.

Are you LLM ready?

Just like traditional SEO relies on clean HTML and structured metadata, LLM Optimization (LLMO) is about giving AI models the most content in the fewest tokens. Right now, llms.txt is a practical first step, helping language models discover and better understand your content, especially technical documentation. It’s the beginning of making your site competitive in AI-driven searches. Google’s search went through many evolutions, and it’s exciting to watch this new wave unfold as AI use grows and better tools emerge.

Are you convinced that AI search optimization is worth doing? Let me know in the comments! 👇🏻