What Is llms.txt, and Can It Really Help Revenue?

What llms.txt is

llms.txt is a proposed convention, published at llmstxt.org in September 2024, for a plain Markdown file placed at the root of a site, at /llms.txt. It gives a short description of the site and a curated list of links to its most useful pages, grouped under headings, so a language model or AI tool can find the important material without crawling everything.

The format is simple: a title, a one-paragraph summary, optional notes, and sections of links with a short description for each. Some sites also publish a longer companion file containing full page text. Think of it as a curated table of contents written for machines, not as a technical control.

What it does not do, and why that matters

Many questions about llms.txt assume it boosts AI visibility or revenue directly. Currently, it does not. No major AI engine has confirmed that it uses llms.txt to decide what to cite or recommend, and Google has said publicly that Search does not use it. Google's guidance for AI Overviews and AI Mode is that no special file or markup is needed; the usual Search fundamentals apply.

It also does not control access. Whether AI crawlers can fetch your pages is decided by robots.txt and by bot settings at your CDN or firewall. A site with a perfect llms.txt and a firewall that blocks OAI-SearchBot will not be cited by ChatGPT search, however well the file is written.

The confusion is understandable. The name echoes robots.txt, and some early commentary described it as a way to steer AI. But the two files do different jobs: robots.txt states crawling rules, while llms.txt is a summary that tools may or may not read.

The honest revenue case

So where could revenue come from? Mostly indirectly. Some AI coding assistants, agents and developer tools read llms.txt when a user points them at a site, which matters for software and API companies whose buyers are developers. Writing the file also forces a useful discipline: deciding which ten to thirty pages best explain what you sell, who it is for and what it costs.

That is the real value: low-cost hygiene that takes an afternoon. Revenue from AI search comes from being named and cited in answers to buying questions, and that depends on access, quotable content, consistent entity data and third-party coverage. Treat llms.txt as a small item on that list, never as the strategy.

There is one more indirect benefit. Writing a clean summary often exposes inconsistencies, such as three different descriptions of the company across the site, or a pricing page that contradicts a comparison page. Fixing those helps every AI engine, whether or not it ever reads llms.txt.

What a worthwhile llms.txt contains

• A one-paragraph summary stating what the company does, for whom and where.
• Links to core commercial pages: products, pricing, comparisons and case studies.
• Links to trust pages: about, licences or certifications, contact and policies.
• Short descriptions that say what each page answers, not marketing slogans.
• Only URLs that are live, indexable and consistent with your sitemap.

Keep it short, accurate and updated when key pages change. A stale file that points to redirected pages does more harm to your credibility with technical buyers than having no file at all.

Worked example: a forex broker's llms.txt

A forex broker might list its account types, a fees and spreads page, its regulation and licence page, platform pages, an Islamic account page and its risk disclosure. Each link gets one plain sentence, for example stating which group entity holds which licence. Because licensing is among the first things buyers ask AI about brokers, the same facts must appear in plain text on the linked pages and match the local regulator's register, so engines find consistent answers wherever they look.

The file will not make ChatGPT recommend the broker. What moves that is accurate, quotable pages, open crawler access and coverage on the comparison sites engines cite.

For a B2B software company the pattern is the same: link product, pricing, integrations, documentation and security pages, and describe each in one factual sentence. For a professional services firm, link practice areas, credentials, case studies and contact details.

Where Suggesting.ai places llms.txt in the work

Suggesting.ai treats llms.txt as hygiene inside a wider programme. The free audit, delivered in 48 hours, checks crawler access including CDN and firewall blocks, live answers on ChatGPT, Perplexity, Google AI and Copilot, competitors named, sources cited, content readiness and entity consistency, and notes whether an llms.txt exists and is accurate.

Organic GEO then focuses on the levers that move answers, and ChatGPT Ads are added where your category and market are eligible. For financial services outside the US, OpenAI's policy generally does not permit ads, and we say so in the audit. The goal is to make AI suggest you, which a text file alone cannot do.

How to tell whether it is doing anything

Check your server logs for requests to /llms.txt and note the user agents. If almost nothing fetches it, that tells you how much weight to give it. Measure revenue impact where it actually shows: AI referrals in analytics, where ChatGPT links carry utm_source=chatgpt.com, mentions and citations in answers to your buyer questions, and leads from those segments.

For a simple test, compare a month of AI referral and citation data before and after publishing the file. Expect little change; if you do see a large one, look for another cause first.