llms.txt Templates by Industry: 6 Copy-Paste Examples for E-commerce, SaaS, Blogs, and More

About this guide

llms.txt is the single most important file for Generative Engine Optimization (GEO), but most people tell me the same thing: "I understand the spec, but I don't know what to actually write for my site." That makes sense — the ideal content of an llms.txt file varies dramatically by the type of site you run.

This guide gives you six industry-specific, copy-paste llms.txt templates: e-commerce, SaaS, blog/media, corporate, documentation, and personal portfolio. Each one is a real starting point — swap in your site name and URLs and you're ready to ship.

If you're still learning the basics, read the llms.txt implementation guide first.

The llms.txt structure (quick recap)

Every llms.txt file uses the same four-part skeleton:

# Site Name

> A one- to two-sentence description (the first thing an AI reads)

## Category name

- [Page title](URL): short description of what's on the page
- [Page title](URL): short description of what's on the page

## Another category

- [Page title](URL): short description

Simple, but AIs use this structure as a map of your site. Sites without a well-organized llms.txt get crawled but rarely cited.

Template 1: E-commerce

Lead with product categories and shopping guides. The goal is for an AI to immediately know what's sold and how to buy it.

# ExampleShop

> Japan-based online store specializing in furniture and interior goods. Free shipping, 30-day returns, and a 3-year warranty on every product.

## Product categories

- [Sofas](https://example.com/category/sofa): 2-seater to 3-seater, leather and fabric
- [Tables](https://example.com/category/table): Dining, coffee, and side tables
- [Storage](https://example.com/category/storage): Chests, bookshelves, shoe racks

## Shopping guide

- [Shipping info](https://example.com/shipping): Rates, delivery times, island surcharges
- [Returns policy](https://example.com/returns): 30-day no-questions-asked returns
- [Warranty](https://example.com/warranty): What our 3-year coverage includes

## Frequently asked

- [Do you offer assembly?](https://example.com/faq#assembly)
- [International shipping?](https://example.com/faq#international)
- [Payment methods](https://example.com/faq#payment)

Key point: Organize by product category, not by individual SKU. Listing every product bloats the file without helping the AI understand your catalog.

Template 2: SaaS (B2B software)

Put features, pricing, and docs front and center so an AI can immediately tell what problem the tool solves.

# ExampleCRM

> Cloud-based CRM for small and mid-size businesses. Contact management, pipeline tracking, email integration, and reporting starting at $20/month.

## Product

- [Features](https://example.com/features): Contacts, pipeline, email integration
- [Pricing](https://example.com/pricing): Starter $20/mo, Enterprise custom
- [Case studies](https://example.com/case-studies): Retail, real estate, manufacturing

## Documentation

- [Getting started](https://example.com/docs/getting-started): 5-minute setup
- [API reference](https://example.com/docs/api): Full REST API v2 reference
- [Integrations](https://example.com/docs/integrations): Slack, Gmail, Google Calendar

## Support

- [Help center](https://example.com/help): FAQ and troubleshooting
- [Contact](https://example.com/contact): Weekdays 9am-6pm

Key point: For SaaS, the documentation section is critical. AIs frequently cite docs when answering technical questions, so making docs easy to discover pays off directly.

Template 3: Blog / media

Lead with categories and flagship posts. The goal is for the AI to recognize what topics you're an authoritative source on.

# Example Tech Blog

> A weekly tech blog for frontend developers, covering React, TypeScript, and performance optimization. Two new posts per week.

## Main categories

- [React](https://example.com/category/react): React 19 features, hooks patterns, architecture
- [TypeScript](https://example.com/category/typescript): Type-safe patterns, advanced type tricks
- [Performance](https://example.com/category/performance): Core Web Vitals, bundle optimization

## Popular posts

- [Complete React Server Components guide](https://example.com/rsc-guide): How RSCs work with real examples
- [Mastering TypeScript type inference](https://example.com/type-inference): 20 practical type tricks
- [7 ways to get LCP under 2.5 seconds](https://example.com/lcp-optimization)

## About

- [Author profile](https://example.com/about): 10 years in frontend engineering
- [Contact](https://example.com/contact)

Key point: The Popular posts section helps AIs prioritize your best content as citation-worthy sources.

Template 4: Corporate site

Company info, products, and press come first. Lead with facts an AI can quote verbatim.

# Example Inc.

> Tokyo-based AI company founded in 2018. 150 employees. Flagship product is the image recognition platform ExampleVision.

## Company

- [About us](https://example.com/about): Founding, leadership, business areas
- [History](https://example.com/history): Key milestones since 2018
- [Team](https://example.com/team): Executive team profiles

## Products & services

- [ExampleVision](https://example.com/products/vision): Image recognition platform
- [ExampleNLP](https://example.com/products/nlp): Japanese-optimized NLP API
- [Case studies](https://example.com/case-studies): Manufacturing and retail deployments

## News & IR

- [Press releases](https://example.com/news): Latest announcements
- [Investor relations](https://example.com/ir): Financial reports, shareholder materials

## Careers

- [Open positions](https://example.com/careers): Engineering, sales, product

Key point: Pack your one- to two-sentence description with the facts an AI will want to quote: headquarters, founding year, headcount, flagship product. It dramatically improves answer quality when someone asks the AI about your company.

Template 5: Documentation / technical site

Hierarchy and completeness matter most. The goal is for an AI to answer technical questions precisely.

# ExampleLib Documentation

> Lightweight HTTP client library for TypeScript. Fetch-based, zero dependencies, fully type-safe.

## Getting started

- [Installation](https://docs.example.com/install): npm / yarn / pnpm
- [Quickstart](https://docs.example.com/quickstart): Learn the basics in 3 minutes

## Core features

- [GET requests](https://docs.example.com/get): Query params and headers
- [POST requests](https://docs.example.com/post): JSON, FormData, multipart
- [Error handling](https://docs.example.com/errors): try/catch and type guards

## Advanced features

- [Interceptors](https://docs.example.com/interceptors): Request/response middleware
- [Retry strategies](https://docs.example.com/retry): Exponential backoff configuration
- [Caching](https://docs.example.com/cache): HTTP cache and custom stores

## Reference

- [API reference](https://docs.example.com/api): All methods and type definitions
- [Migration guide](https://docs.example.com/migration): v1 to v2 upgrade path

Key point: Order sections in the natural learning flow: Getting started → Core → Advanced → Reference. AIs follow this structure when answering tutorial-style questions.

Template 6: Personal portfolio

Keep it simple: profile, work samples, and contact info. Make it instantly clear what kind of work you take on.

# Taro Yamada | Web Designer

> Tokyo-based web designer specializing in corporate sites, landing pages, and branding. 100+ projects delivered over 10 years.

## About me

- [Profile](https://example.com/about): Background, specialties, tools
- [Skills](https://example.com/skills): Design, HTML/CSS, Figma, Webflow

## Work

- [Corporate site projects](https://example.com/works/corporate): 10 featured projects
- [Landing pages](https://example.com/works/lp): 20 LP projects
- [Branding work](https://example.com/works/branding): Logo and guideline design

## Blog

- [Design thinking](https://example.com/blog/design): Essays on web design
- [Project journal](https://example.com/blog/diary): Behind-the-scenes posts

## Contact

- [Contact form](https://example.com/contact): Project inquiries and quotes
- [X (Twitter)](https://x.com/example): DMs open

Key point: For personal sites, the three signals to nail are what you do, proof you can do it, and how to reach you.

Cross-industry best practices

✅ Do

1. Pack the top description with essential facts — site name, location, focus area, unique value
2. Keep categories to 5–7 — more than that and the AI struggles to grasp the structure
3. Add a one-line description to every page — title alone isn't enough
4. Use absolute URLs — never relative paths
5. Stay shallow — H2 with lists is the sweet spot; avoid deep H3/H4 nesting

❌ Don't

1. Don't list every page — llms.txt isn't a sitemap; stick to key pages
2. Don't use marketing hyperbole — "industry leader," "the only," and similar phrases get flagged
3. Don't let it go stale — outdated llms.txt signals a dead site
4. Don't skip llms-full.txt — if you have bandwidth, add a detailed companion file
5. Don't forget to verify — load /llms.txt in a browser after deployment

How to install llms.txt

Once you have your file ready:

1. File name must be exactly llms.txt (lowercase, plural)
2. Place at your domain root: https://yourdomain.com/llms.txt
3. Serve with Content-Type: text/plain; charset=utf-8
4. Confirm robots.txt isn't disallowing it
5. Hit yourdomain.com/llms.txt in a browser to verify it's live

FAQ

How is llms.txt different from sitemap.xml?

sitemap.xml provides a machine-readable list of all page URLs for search engines. llms.txt is a human-readable summary of your site's main pages that helps an AI understand who you are. Ideally you publish both.

Can I use these templates as-is?

Swap in your own site name, URLs, and descriptions before publishing. AIs detect duplicate patterns, so identical llms.txt files across multiple sites could hurt your credibility signals.

I published llms.txt but I'm still not cited by AI — why?

llms.txt is just the key that unlocks the door. Also check whether GPTBot, ClaudeBot, and PerplexityBot are allowed in your robots.txt, and whether your structured data and content quality are in shape. You can verify all 11 GEO items at once with IndexReady.

Audit your llms.txt for free

Whether your llms.txt is deployed correctly and your site is otherwise AI-ready can be checked in seconds at IndexReady. Enter any URL to get a 100-point GEO score covering llms.txt, AI crawler permissions, structured data, and more.

If you need the basics first, start with the llms.txt implementation guide.