Run by Claude

AI for Small Business

You Rank on Google. ChatGPT Has Never Heard of You.

Most small businesses have spent years winning at Google SEO. ChatGPT runs a different game — with different signals, different rankings, and a different winner.

MurphMay 22, 20265 min read

I've been watching a pattern emerge across every audit we run.

Business has great Google rankings. Has been at it for years. Shows up consistently in local search. Owner is proud of it, and they should be — that work took real time and money.

Then I ask: "What happens when someone asks ChatGPT who the best [service] is in [city]?"

They've never checked.

We check it together during the audit. Maybe three times out of ten is the business actually in the response. The other seven: a competitor comes up, a generic answer, or the AI hedges and says it doesn't have specific local information.

That's the Google vs. AI gap. It's real, it's growing, and most business owners don't know it exists.


Two Different Games

Google built its index over 25 years. The rules are known: keywords, backlinks, page authority, technical SEO. Play the game well and you rank. The signals are clear.

AI assistants run a different game. ChatGPT, Claude, Perplexity, Gemini — they don't use Google's index. They use:

  • Training data — everything absorbed during training. If your business wasn't well-documented on the web during that window, you're a gap.
  • Real-time crawling — AI systems with web access can read your site right now. What they find, and how well they understand it, determines what they say about you.
  • Structured signals — schema markup, llms.txt, FAQ structured data. Machine-readable formats that tell AI systems exactly what your business does without ambiguity.
  • Third-party citations — directories, reviews, news mentions, industry listings. Every external reference to your business reinforces your signal.

Here's the thing: you can win one game completely and be a ghost in the other.

A business with strong Google SEO might have zero schema markup (because old-school SEO didn't require it), no llms.txt (brand new standard), thin FAQ content, and robots.txt accidentally blocking AI crawlers.

All four of those are AI visibility failures. None of them affect Google rankings.


What AI Systems Actually See

When someone asks ChatGPT "who's the best HVAC contractor in Columbus?", the AI does something like this:

  1. Searches its training data for mentions of HVAC contractors in Columbus
  2. If it has real-time web access, runs a quick search
  3. Checks what structured data it can parse from company websites
  4. Weighs third-party credibility signals (directories, reviews, citations)
  5. Synthesizes a response

If your HVAC business has strong Google rankings but no schema markup, no llms.txt, blocks AI crawlers in robots.txt, and hasn't been mentioned in recent web content — you might as well not exist to that system.

Meanwhile, a competitor with slightly worse Google rankings but solid structured data, an llms.txt file, and active content creation might be the AI's first recommendation.


The 5 Signals That Determine AI Visibility

Here's what we check in every audit:

1. llms.txt file — The structured business summary at your domain root. New standard. Less than 5% of small businesses have one. The AI equivalent of a Google Business Profile — except most of your competitors haven't built it yet.

2. LocalBusiness schema — JSON-LD markup that explicitly identifies your business type, location, hours, services, and contact info. The highest-signal structured data you can add.

3. FAQ schema — Pre-formatted question-and-answer pairs that AI systems can pull directly into responses. "How much does gutter cleaning cost?" — if you have FAQ schema answering this, you're the answer.

4. AI crawler access — Is your robots.txt blocking GPTBot, ClaudeBot, or PerplexityBot? Many sites do this accidentally through default CMS configurations. If AI crawlers can't read your site, they can't recommend you.

5. LLM knowledge test — We ask a language model directly about your business. This baseline tells you what AI assistants currently know, what they get wrong, and what gaps to fill first.


Why the Window Is Now

Most of your competitors are still in Google-only mode. They're optimizing title tags and chasing backlinks — which still matters — but they haven't thought about the AI layer at all.

That creates a window. The businesses that build AI visibility in 2026 will own it by 2027. The ones that wait until everyone's doing it will be paying to catch up, the same way businesses that waited on Google SEO until 2015 had to fight for ground that early movers owned for free.

The free audit at /start takes two minutes. It checks all five signals, shows how your top competitor compares, and gives you a prioritized fix list.

If you're not in the AI answer when your customers ask, now is the time to fix it.

Run the free audit →


— Murph, VibeTokens

Want to see how your business stacks up?

Get a free brand audit — we'll show you what's working, what's not, and what to fix first.

Free Brand Audit →

Frequently Asked

Why does my business rank on Google but not appear in ChatGPT or Claude?

Google and AI assistants use completely different systems. Google uses keywords, backlinks, and page authority. AI systems pull from training data (which may predate your business or have gaps), real-time web crawling, and structured signals like schema markup and llms.txt. A #1 Google ranking doesn't translate to AI visibility unless you've built the specific signals AI systems look for.

How can I check if AI assistants know my business?

Open ChatGPT, Claude, or Perplexity and ask: 'Who are the best [your service] businesses in [your city]?' Then ask: 'What do you know about [your business name]?' The answers tell you exactly where you stand. Our free audit at vibetokens.io/start runs this check automatically and shows you how to fix any gaps.

How long does it take for AI assistants to start recommending my business after I fix my AI visibility?

Adding llms.txt and schema markup can affect AI systems with real-time web access (Perplexity, SearchGPT) within days. For AI systems that rely on training data, you're building signals for the next training cycle — typically months out. The window to get ahead of competitors is wide open right now.

Jason Murphy

Written by

Murph

Jason Matthew Murphy. Twenty years building digital systems for businesses. Former CardinalCommerce (acquired by Visa). Now running VibeTokens — a brand agency for small businesses that builds websites, content, and growth systems with AI.

Live Workshop · April 27

Build your Claude OS in 4 hours. CLAUDE.md, MCP servers, 3 custom workflows. 8 seats, $247.

Reserve Seat →

Your brand is your first impression.

Find out if it's costing you customers.

Free brand audit. We analyze your online presence, competitors, and messaging — then tell you exactly what to fix.

Get Your Free Brand Audit →