How to Get Your Business Cited by ChatGPT, Perplexity, and Google AI Search

The question has changed.

It used to be: how do I rank on Google? Now it's also: how do I get ChatGPT to mention my business? These are related problems but not identical ones. Ranking in traditional search and getting cited in AI-generated answers require some of the same work, and some that most businesses haven't done yet.

This guide covers both. Everything here is what Working Model has done for clients and for our own site. No theory.

How AI search engines find and cite businesses

The mechanics matter, because they determine what's worth doing.

AI answer engines (ChatGPT with search, Perplexity, Google AI Overviews) don't rank websites the way traditional search does. They synthesize answers from sources they've determined to be credible and relevant, then cite those sources. The question they're asking about your site isn't "does this page rank for this keyword?" It's "can I trust this source to answer this question accurately?"

That means they look for: clearly structured content that directly answers questions, consistent entity data across multiple sources, structured data that signals what your business does, and third-party corroboration that your business is what it claims to be.

You can influence all of these.

Step 1: Get your technical foundation right

AI systems can't cite what they can't read.

Before anything else: verify your site is actually accessible to crawlers. This sounds obvious. It's surprising how often it's wrong.

• No accidental noindex tags on pages you want indexed
• Clean robots.txt: nothing blocking crawlers that should have access
• Pages load fast (Core Web Vitals matter; slow sites get deprioritized)
• No broken canonical tags pointing pages at the wrong URL
• Sitemap is current and submitted to Google Search Console

If you're not sure, Google Search Console's coverage report will show you what Google can and can't see. Fix what's broken there first. Everything else depends on this.

Step 2: Add structured data

Schema markup is one of the clearest signals available to AI systems. It doesn't just help Google understand your pages. It tells any system reading your site exactly what your business is, what it does, and where it operates. Plain-language content leaves room for interpretation. Schema doesn't.

Minimum viable schema for a service business:

• Organization: your business name, URL, description, and sameAs links to your social profiles and directories
• LocalBusiness or ProfessionalService if your location matters to clients
• Service on service pages: what the service is, who it's for, where it's offered
• Article on any content you publish

The sameAs field in Organization schema is underused and undervalued. It explicitly connects your website to your Google Business Profile, LinkedIn page, Clutch listing, and anywhere else you appear online. That's the structured version of entity consistency.

Schema.org has documentation for all of these types. Google's Rich Results Test will tell you if your markup is valid.

Step 3: Create an llm.txt file

Most businesses have never heard of this. It's one of the most direct things you can do.

An llm.txt file is a plain-text file that lives at your domain root (yourdomain.com/llm.txt) and gives AI crawlers a structured, accurate summary of what your business does, who it serves, and what content exists on the site. Think of it as a robots.txt for meaning rather than access.

What to include: a brief, factual description of your business; your services with plain-language descriptions; your key pages with URLs; contact information. Write it the way you'd describe your business to someone who'd never heard of it. No marketing language. AI systems aren't moved by it, and it wastes space that could be used for useful signal.

Working Model's llm.txt is public. It's a straightforward example of the format.

Step 4: Publish content that directly answers questions

This is where most business websites fail, and where the opportunity is largest.

AI answer engines quote sources that clearly and directly answer the question being asked. A service page written in marketing language ("we deliver transformative results for growth-focused organizations") is nearly uncitable. A page that asks "what does a brand strategy engagement cost?" and answers it in the first paragraph is highly citable.

The structural requirement is simple: clear question as a heading, direct answer in the first two or three sentences below it. Everything after that is elaboration.

In practice: think about the questions your clients ask in the first meeting. Write a direct answer to each one. Publish them as blog posts, FAQ sections, or standalone pages. Five honest answers to real questions will do more for AI citability than a perfectly written services page.

Step 5: Build consistent entity presence across the web

AI systems cross-reference multiple sources to verify that your business is what it claims to be. If your business name, address, service descriptions, or positioning read differently on your website than on your Google Business Profile than on Clutch. The system has conflicting signals. Conflicting signals reduce confidence. Reduced confidence means less citation.

The sources that matter most:

• Google Business Profile: name, address, phone, and services must match your website exactly
• Clutch or equivalent industry directory (Semrush Agency Directory, G2, whatever's relevant to your category)
• LinkedIn company page: description, services, and size should be consistent
• GitHub: if your business produces any public code or tools, that's a credibility signal in technical categories. Working Model's open-source work registers here.
• Any publication mentions: even local press counts

The goal isn't maximum volume of listings. It's consistency across the listings you have.

Step 6: Get third-party validation

LLMs are skeptical of self-citation. A business that only talks about itself is harder to cite with confidence than one that others talk about.

The practical version of this:

• Ask current and past clients for Google reviews. Five honest reviews from real clients make a measurable difference. Fifty generic five-star reviews with no text make almost none.
• Pursue one or two mentions in publications your audience actually reads. A single article in a credible industry publication outweighs dozens of directory listings.
• If any of your clients publish case studies or testimonials on their own sites that name you. That's the strongest signal of all. It's independent corroboration from a source with its own domain authority.

You can't manufacture this quickly. But you can stop leaving it on the table.

Step 7: Allow AI crawlers explicitly

This one surprises people. Many businesses are accidentally blocking the crawlers that AI search systems send to index content.

Check your robots.txt. Explicitly allow: GPTBot, ChatGPT-User, CCBot, anthropic-ai, Claude-Web, PerplexityBot. Some security tools and CDN configurations block unfamiliar bots by default. AI crawlers often get caught in these.

Working Model's robots.txt shows the right setup. If your file doesn't explicitly allow these agents, AI systems may not be able to crawl your site at all, which makes everything else in this guide moot.

How long does this take?

Honest answer: it depends on the type of change.

Structural changes (schema markup, llm.txt, robots.txt) have near-immediate effect on crawlability. Once a crawler can read your site correctly and understand what your business is, that signal is available on the next crawl.

Content and entity signals take longer. Search integrations and AI training pipelines update on their own schedules. Realistically, expect 4–12 weeks before content changes show up in AI-generated answers.

Third-party signals take as long as they take. There's no shortcut to a genuine Google review or a press mention. The businesses that treat this as ongoing practice rather than a one-time campaign compound the advantage over time.