You don't show up in AI search answers
measuring ChatGPT, Perplexity, Gemini and AI Overviews
The agent measures your brand's visibility across AI search engines, checks whether models recognise you correctly, and fixes fixable gaps with a pull request.
This is you if
- Customers increasingly ask AI search engines, not Google.
- You don't know whether ChatGPT or Perplexity mention you at all.
- When a model does mention you, it confuses your company with someone else.
- Competitors show up in AI answers and you don't.
What it looks like today
You'll recognise it from a few things no one measures today.
- No one in the company knows how often AI models cite you.
- Your site is missing structured data and llms.txt.
- Wikidata and other sources about your company are incomplete or missing.
- You never find out when models describe you incorrectly.
- Someone occasionally types a query into ChatGPT by hand and watches what comes out.
- The result is never recorded and can't be compared over time.
- Structured data goes untouched because it's unclear what helps.
- Errors in how models identify you stay unfixed.
- The agent tests your brand across models on schedule.
- It measures citation rates over time, so you see the trend.
- Schema and content fixes arrive as pull requests.
- An entity audit shows where sources are missing data.
The agent runs in the background so AI search engines know and cite you, measuring whether it improves.
- 01
Visibility testing
Tests your brand across ChatGPT, Gemini and Perplexity with intent-segmented queries and tracks citation rates over time.
- 02
Entity & knowledge audit
Checks whether models identify you correctly and audits Wikidata, Knowledge Panel and Crunchbase.
- 03
Optimization for AI
Scores pages for AI citability, audits JSON-LD schemas and tunes llms.txt.
- 04
Automated fixes
Generates schema, FAQ sections and llms.txt improvements as pull requests and measures impact before and after.
The agent has 22 tools across six categories, runs daily, weekly and monthly, and tests your brand across multiple models at once. Trend reports go to Discord.
I run it for cyrcID and it's open-source on GitHub.
What you get
- An agent that measures your brand's visibility in AI search.
- An audit of whether models recognise you and where data is missing.
- Schema and content fixes as pull requests to approve.
- Operation and measurement of whether citations grow over time.
How it runs after launch
Launch isn't the end. I run and maintain the agent for you and track whether AI search cites you more.
More on how I handle security and reliability.
01How do you even measure visibility in AI search?
The agent sends intent-based queries to the models and tracks whether and how they mention you. It stores the results, so you see citation rates over time, not one random attempt.
02Isn't regular SEO enough for this?
SEO is about rankings in Google, but AI models assemble an answer differently and cite different signals. This agent targets whether models know and mention you, which classic SEO doesn't measure.
03What does it connect to?
Your site for schema and llms.txt fixes, and public sources like Wikidata. Fixes come through normal Git for approval.
04How long does deployment take?
I usually have the first visibility measurements running in days. You get a precise estimate after a short look at where your brand stands.
05What does it cost?
Running the agent is a monthly fee, not a bill for hours. The exact figure depends on scope, no blind pricing.
How to find out whether AI search engines even know you.
Eight steps to check whether ChatGPT, Perplexity and the rest mention you, and what to do when they don't. I'll send them to your inbox.
- How to ask the AI engines yourself and see whether they cite you at all.
- Which pages you need so models tie you to your field.
- Why ordinary SEO isn't enough and what AI engines look for on top.
- How to get into the sources models pull their answers from.
- What to track so you spot your AI visibility before competitors do.