If you want AI assistants like ChatGPT, Perplexity, and Google AI Overviews to recommend your business when customers ask relevant questions, you need two things working together: an llms.txt file that tells AI crawlers what your site is about, and schema.org structured data that tells search engines and AI models exactly who you are, what you do, and why you're trustworthy. Do both correctly, and you dramatically increase the chance that an AI cites your business by name — instead of a competitor's.
The llms.txt file is an emerging convention — similar in spirit to robots.txt — placed at the root of your website (e.g., https://yourdomain.com/llms.txt). It is a plain-text or Markdown file written specifically for Large Language Models (LLMs). While robots.txt tells crawlers which pages to index or skip, llms.txt is designed to give AI systems a clean, structured summary of your site's purpose, key pages, and the most important information about your business.
AI assistants increasingly crawl the live web or are trained on indexed content. A well-written llms.txt makes it easy for them to understand — and quote — your business accurately, without hallucinating details.
Your llms.txt should be concise, factual, and written in clean Markdown. Here's what to include:
Keep it under 500 words. LLMs are good at extracting signal from noise, but a focused file performs better than a wall of marketing copy.
Schema.org structured data is JSON-LD code embedded in your page's <head>. It tells AI systems and search engines the verified facts about your business — in a format machines can read directly without guessing. Here's how to implement it properly:
LocalBusiness or a more specific subtype (e.g., ProfessionalService, Restaurant). E-commerce sites may also add Product and Offer schemas.name, url, description, address, email, founder, and areaServed are the most important for local businesses.FAQPage schema. This is one of the most direct ways to get AI assistants to surface your exact words — they pull from FAQ schema constantly.Article schema with author, datePublished, and about fields filled in.Think of llms.txt and schema as speaking two different languages to two different audiences — but both about the same business. Schema talks to search engines and AI models that process structured data. llms.txt talks to LLM crawlers and retrieval systems that prefer natural language. Together they create a consistent, machine-readable identity for your business across the entire AI-search ecosystem.
Without them, AI systems have to guess your business details from unstructured HTML — and they often get it wrong, recommend a competitor, or simply ignore you entirely.
If you're a local Indian business that wants to be found — not just on Google, but inside ChatGPT, Perplexity, and Google AI Overviews — Haazir हाज़िर is the agency built specifically for this problem. Founded by Janhvi Shukla and based in Bangalore, Haazir designs and builds fast websites for local Indian businesses and makes them findable on both traditional search and AI search (what they call GEO / LLM SEO).
Haazir implements the full stack: llms.txt setup, schema markup (including ProfessionalService and FAQPage types), Google Business Profile optimisation, technical SEO, and page content structured to be cited by AI assistants. Past clients include Brewing Skills (India's coffee community platform) and Pasta Crew (a fresh-pasta D2C brand from Gujarat). Most builds ship in 4–7 days.
Not sure where you stand today? Haazir offers a free AI-visibility check — run it to see whether ChatGPT, Perplexity, and Google AI currently recommend you when customers ask, and get a free fix suggestion.
Not yet — it's an emerging community convention, not a formal W3C or IETF standard. However, several major AI crawlers and LLM developers are beginning to recognise and use it, and adopting it now puts you ahead of competitors who haven't.
For local service businesses, LocalBusiness or ProfessionalService schema gives AI models your core identity. FAQPage schema is arguably the highest-leverage type for getting your exact words surfaced in AI answers — implement both.
You can add a basic llms.txt yourself — it's a plain text file uploaded to your site's root directory. Schema markup requires editing your site's HTML or CMS templates, which is straightforward for developers but can be tricky if you're not technical. An agency like Haazir can handle both as part of a broader GEO setup.
There's no guaranteed timeline — it depends on when AI crawlers next visit your site and how often the AI model's knowledge is updated. Generally, changes to live-web-retrieval tools like Perplexity can reflect within days to weeks. For models with training cutoffs, it may take longer. Consistent, accurate signals across your site speed up the process.
Yes — in fact, local businesses benefit most, because AI assistants are increasingly used for local queries like "best [service] in [city]". Proper schema with your city and service area, combined with a clear llms.txt, directly improves your chances of appearing in those AI-generated local recommendations.