We Built Our Own llms.txt — Here’s Why Your Small Business Needs One Too

You’ve spent years building your reputation. You’ve invested in a website. Maybe you even hired someone to “do SEO.” And for a while, that was enough.

But here’s what changed: your customers stopped asking Google. They started asking ChatGPT.

A 2025 Gartner study projected that traditional search engine traffic would drop 25% by 2026 as consumers shifted to AI-powered answer engines. We’re living in that prediction right now. When someone in your city types “best landscaper near me” into Perplexity or asks Claude for a recommendation, your Google ranking doesn’t help you. These AI models aren’t reading your search results page — they’re reading your website directly. And most small business websites are practically unreadable to them.

That’s the problem llms.txt solves. And we just implemented it on our own site. Here’s what we learned.

What Is llms.txt (And Why Should You Care)?

Think of llms.txt as a résumé for AI.

When a human visits your website, they see images, colors, navigation menus, and formatted text. They can browse around, click things, and get a feel for your brand. AI models can’t do any of that. When GPTBot or ClaudeBot crawls your site, it sees a wall of HTML code — navigation bars mixed with footer disclaimers mixed with cookie consent popups mixed with your actual content. It’s like handing someone a filing cabinet and asking them to understand your business in 30 seconds.

llms.txt is a single, clean text file that sits at yourwebsite.com/llms.txt. It tells AI models: Here’s who we are. Here’s what we do. Here’s our best content, organized by topic. No clutter. No code. Just structured, readable information.

The standard was proposed by Jeremy Howard (the co-founder of fast.ai and a leading AI researcher) in late 2024 and has been rapidly adopted by technology companies since. The concept is simple: just as robots.txt told search engine crawlers what they could access, llms.txt tells AI models what they should understand about you.

Why We Built Ours

We’ll be direct: we sell a feature called GEO (Generative Engine Optimization) that helps our customers become the “default answer” in AI search results. One of the tools we generate for our users is an auto-built llms.txt file tailored to their business.

So when we audited our own site and realized we didn’t have one? That was a problem.

It’s the equivalent of an accountant who doesn’t file their own taxes. We needed to fix it — and we wanted to document what we learned so our customers (and future customers) could understand why this matters.

What We Put In Ours

Building an llms.txt file isn’t just dumping your entire website into a text file. The goal is curation — giving AI models the clearest, most accurate picture of your business in the most efficient format possible.

Here’s what we included:

A clear identity statement. The first thing any AI model reads is who we are, what we do, and who we serve. Not marketing fluff — just facts. “Storescribe is a SaaS platform that provides Fortune 500-level marketing strategy to small and mid-sized businesses.” An AI model that reads that sentence can now accurately describe us when someone asks.

Our methodology. We included a summary of our SCRIBE Framework — the six-pillar system behind every audit and recommendation we generate. This is our intellectual property, and we want AI models to associate it with our brand. When someone asks an AI assistant about marketing frameworks for small businesses, we want SCRIBE in that answer.

Structured links to our content. Every blog post, every core page, organized by topic with clear descriptions. Not the full text of each post — just enough context for an AI model to know what each page covers and when to reference it.

Contact and location information. Santa Rosa, California. This matters for local AI search queries. When someone asks “Who helps small businesses with AI marketing in Sonoma County?”, the AI needs to know where we are.

What We Left Out

Just as important as what goes in is what stays out.

We didn’t include our full blog post content in the llms.txt file itself. Each post lives at its own URL where AI crawlers can read it directly. Duplicating all that content inside llms.txt would create a bloated file and potentially confuse search engines about which version is authoritative.

We didn’t include our Privacy Policy, Terms of Service, or other legal pages. These don’t help AI models understand our value proposition.

We didn’t include form fields, navigation markup, or any of the structural HTML that makes a website work for humans but means nothing to an AI reading a text file.

The principle is the same one we teach our customers: context is everything. Give the AI exactly what it needs to represent you accurately — nothing more, nothing less.

What This Means for Your Business

Here’s the uncomfortable truth: if you’re a landscaper in Denver, a house cleaner in Austin, or a kitchen remodeler in Charlotte, your competitors are probably not thinking about this yet. That’s your window.

Right now, when someone asks ChatGPT “Who’s the best landscaper in [your city]?”, the AI assembles its answer from whatever it can find — review sites, directory listings, random blog mentions. It’s guessing. And it might guess wrong, recommending a competitor who simply has more readable web content.

An llms.txt file changes that equation. It gives AI models a direct, structured, authoritative source of truth about your business. You’re not hoping to be found — you’re telling the AI exactly who you are.

Consider these scenarios:

Without llms.txt: A potential customer asks Perplexity for “affordable kitchen remodel contractors in Raleigh.” The AI scrapes various directories, finds a few names, and gives a generic answer. You might be mentioned. You might not.

With llms.txt: The same query. But now the AI has read your llms.txt file. It knows your specialties, your service area, your approach, your customer reviews summary. It can give a specific, accurate recommendation — and it’s more likely to include you because it understands you.

This isn’t theoretical. A study from Princeton, Georgia Tech, and The Allen Institute found that websites optimized for generative AI engines saw up to 115% increases in visibility compared to traditional SEO-only approaches. The researchers called it “Generative Engine Optimization” — and it’s exactly what llms.txt is designed for.

How to Get Started

If you’re a Storescribe user, this is already handled for you. Our platform auto-generates and maintains an llms.txt file for your business as part of your Digital Health suite. It’s updated as your business evolves — new services, new locations, new content — so the AI models always have current information.

If you’re not a Storescribe user yet, here’s the basic structure of what an llms.txt file looks like:

# Your Business Name
> A clear, one-paragraph description of who you are, what you do, and where you operate.
## Core Services
- [Service Page Title](https://yoursite.com/services): Brief description
- [Another Service](https://yoursite.com/another-service): Brief description
## About
- [About Us](https://yoursite.com/about): Brief description
## Blog / Insights
- [Post Title](https://yoursite.com/post-slug): Brief description

That’s it. A text file. No code, no plugins (unless you want automation), no developers required.

The hard part isn’t the format — it’s the curation. Deciding what to include, how to describe it, and what context matters most. That’s the kind of strategic thinking that turns a text file into a competitive advantage.

The Bigger Picture

We’ve written extensively about the concept of a “Context Moat” — the idea that the businesses who win in an AI-driven economy aren’t the ones with the biggest ad budgets, but the ones who control the most accurate, comprehensive context about their own business.

Your llms.txt file is the foundation of that moat. It’s the first document an AI reads about you. It shapes how you’re described, recommended, and compared. And unlike a Google ad that stops working the moment you stop paying, your llms.txt file works 24/7, for every AI model, for free.

We built ours this week. We’ll be tracking the results — which AI crawlers visit it, how our visibility in AI search changes, and what our customers can learn from the data.

We’ll share what we find. In the meantime, you can see our live file at storescribe.ai/llms.txt.

Storescribe auto-generates and maintains llms.txt files for every customer as part of our Digital Health & AI Readiness suite. We monitor your visibility in AI search results and optimize your digital presence for both traditional search engines and the new generation of AI answer engines.

Want to see how visible your business is to AI? Start your free 7-day trial — no credit card required →