SEO for Google AI Mode: what to focus on instead of llms.txt
Google has published its official guide for AI Overviews and AI Mode. Here is where a small business should spend its budget, in plain words.
3 min read
The question I have heard most in recent months is this: "Do we need to prepare our site for AI search?" Google announced its first comprehensive official guide on the subject on 15 May 2026. I read it and wrote this summary for business owners, because some fairly expensive offers are going around on this topic.
What Google says
According to the Google Search Central guide, generative AI results such as AI Overviews and AI Mode rely on Search's core ranking and quality systems. That means the SEO basics we already know still apply. The points I found most relevant for businesses:
- Google Search does not use llms.txt. The file has no positive or negative effect on your visibility or ranking in Google.
- No special schema markup is needed for AI search. There is also no need to chop content into artificial chunks or rewrite it just for AI.
- Google recommends original content based on the business's own experience, with expert input, over generic text that could appear anywhere.
- It recommends tracking visibility with the Generative AI performance report in Search Console.
What this means for your budget
If someone tries to sell you an llms.txt file or a special "AI schema" setup as "AI optimisation", know that it does nothing in Google. Other AI services may use llms.txt; Google's statement only covers Google Search. Before spending on that kind of work, I think the sensible move is to ask the provider for concrete proof that it is used.
The money does real work on the basics your site is already missing. If your service page only says "we provide quality service", the original content Google recommends is not there. A real project example, how the work runs, what affects the price and your experience in your own city are useful to visitors and to search systems alike.
What a concrete service page looks like
Picture an architecture office in Erzurum. A weak service page says "we create modern and functional projects" and stops there. The page we want shows a few projects the office finished in the last two years, with photos, explains the steps a project goes through from the first meeting to the permit, describes how floor area, the soil survey or the number of revisions affects the price, and talks about how the city's climate shapes the design. None of this can be copied from another site, because it comes from the office's own experience. To me that is exactly the content Google describes in its guide.
You don't need an agency or an AI tool to write it. Usually it is enough to put into writing what you already tell customers on the phone. On my projects I often collect this in a half-hour conversation with the owner and then put it on the page.
The steps I recommend
- Add real project examples, process details, the factors that affect price and your local experience to your service pages.
- Check that your business details, product data, images and videos are crawlable and up to date.
- Fix technical SEO problems: internal links, indexability and page experience.
- If you can see the generative AI performance report in Search Console, look at query and page changes once a month.
- Ask for proof before putting budget into llms.txt or a special "AI schema" project aimed only at Google.
Limits worth knowing
AI features and Search Console reports may not appear the same way in every country or account. If you cannot see the report yet, that does not mean something is wrong with your site. I would also add that this guide was written for Google's own product. Other AI tools may look at other sources, but for most small businesses in Turkey and Europe, traffic still comes largely from Google.
If you would like to go through where your site stands on these basics, write to me.
Sources
- A new resource for optimizing for generative AI in Google Search, Google Search Central, 15 May 2026
- Optimizing your website for generative AI features on Google Search, Google Search Central, last updated 10 July 2026