Contents

If your website brings in calls and inquiries, you probably want to show up wherever customers search. Lately, that includes AI tools.
One idea getting a lot of attention is a small file called llms.txt. Some marketers pitch it as an easy way to get noticed by ChatGPT and Google’s AI features.
However, the available evidence points in a different direction. Since time and budget are limited, your effort may go further when you make pages easier for people and machines to read and use.

llms.txt is a plain text file that usually sits at the root of a website. It lists important pages with short descriptions, much like a table of contents for AI systems.
The file is still a proposal rather than an established standard. Unlike robots.txt, which tells crawlers where they may go, it does not control access.
Listing a page in llms.txt also does not require any AI system to read or cite that page.

In mid-2026, Google updated its guidance on its AI search features. It explains that site owners do not need special AI text files or extra markup to appear in Search, including its AI results.
Google also states that Search ignores llms.txt. As a result, keeping one neither helps nor hurts rankings there.
In Google’s view, optimizing for AI search is still SEO at its core.
Independent data points the same way. In an Ahrefs study of roughly 137,000 domains, the vast majority of llms.txt files received no requests in May 2026.
A separate SE Ranking analysis looked at 300,000 domains. It found no measurable link between having the file and being cited by AI.
Other AI companies have not stressed the file either. OpenAI, Anthropic, and Perplexity publish guidance for site owners, but that guidance focuses on crawler access.
A Google search representative has compared llms.txt to the old keywords meta tag. Both rely on a site describing itself, even though a system could simply check the site directly.

Even so, the file does seem to have a real audience in AI coding assistants. These tools may use it to find their way through developer documentation, usually after someone points them to it.
That use case is unlikely to apply to a local service business or small retailer. For most business owners, llms.txt is a lower priority than improving the site itself.

Accessibility means building pages that more people can use comfortably. That includes people who rely on screen readers or keyboards.
Several basics support this goal, such as clear contrast, readable text, image descriptions, and logical headings. They make content easier to follow, and they can help automated systems understand how a page is organized.
That said, there is little evidence that accessibility works as a standalone ranking boost. Instead, it supports the relevance and quality signals that search systems already weigh.
Google’s Lighthouse is a diagnostic tool for testing websites. It now includes checks for how ready a site is for AI agents, which are tools that browse websites for people.
These checks review the same labels and page structure that screen readers use. They also flag layouts that shift unexpectedly.
It is worth noting that these checks measure agent readiness, not search rankings. Still, they suggest that clean page code, stable layouts, and well-labeled content may matter to visitors and AI tools alike.

Experienced SEO and web design teams tend to strengthen the foundation before adding optional extras. The exact priorities vary by site, but the work often covers these areas.
This order has practical value for businesses that depend on inquiries. When visitors can read, compare, and act without fighting the page, they may be more likely to remember the business and return.

llms.txt may serve a narrow purpose, but current evidence does not support it as a path to better rankings. Meanwhile, the work that helps real visitors use a website appears to be what search engines and AI tools rely on too.
If your goal is to turn more qualified visitors into customers, accessible and well-organized pages are a reasonable place to focus.
Usually not. Lighthouse treats a missing llms.txt file as not applicable, so it is not counted as a failure.
It may be worth considering if your site publishes developer documentation. It may also make sense if an AI platform that sends you customers asks for the file.
In most cases, there is no need. Google says keeping the file for other systems is fine, and it does not affect Google Search either way.
A short list of key pages with accurate descriptions tends to work better than a long one. Reviewing the list when those pages change helps keep it current.
Diagnostic tools such as Lighthouse can flag some common issues. From there, a review of headings, contrast, image descriptions, and mobile use is a practical next step.
Have any questions or comments? Write them below!