CitedBy: AI visibility audit

Does ChatGPT recommend your business or your competitors?

We ask ChatGPT, Perplexity and others real customer questions and show you how visible your business is in AI answers.

audit/newFree
Why this matters

Your customers no longer compare ten results. They ask once.

Anyone looking for a law firm, a practice or a tradesperson today gets no list of results from ChatGPT, Perplexity or Gemini, but an answer with a few names. A business that is not in it does not exist for that enquiry.

01

There is no second page

An AI answer names a handful of businesses and then it ends. There is no place further down: your name is in the answer or it is not.

02

It never shows up in your numbers

Someone who gets your competitor's recommendation never visits your website. That lost enquiry appears in no website statistics: there is neither a click nor a bounce.

03

A good Google position is not enough

AI models answer from training data and, depending on the model, from their own web search. Your position in the classic results list does not decide whether you get named.

Method

A measurement record, not a snapshot.

You can ask ChatGPT about your industry yourself. But a single answer is chance: different wording, different model, different day, different result. That is why we measure systematically.

measurement method
  1. 1

    Generate buying-intent prompts

    From your industry, city and country we create up to 30 questions the way real customers ask them: from "Best employment lawyer in Leeds?" to "Top 5, please list them by name". The questions are about the service, never your name. Otherwise the question would already contain the answer.

  2. 2

    Ask several AI models

    Each question goes to several models, some of them with live web search. The number of questions and models depends on the plan: the free audit is a sample, the paid plans measure more broadly.

  3. 3

    Evaluate the answers

    The recommended businesses and their order are extracted from each answer. Spelling variants of your name (legal form, accents, short form) are merged so that no mention slips through.

  4. 4

    Calculate the score

    Mention rate (70 percent) and position within the answer (30 percent) give a value between 0 and 100. First place counts fully, each further position proportionally less.

Limits of the measurement

AI answers are not deterministic: the same question can be answered differently tomorrow. The score is therefore a trend across many queries, not an exact single measurement.

It gains meaning through comparison: against the businesses named instead of you, and against your own value from last week. Nobody can promise a guaranteed placement in AI answers, and neither can we.

We ask in German or English and name the country in the question. The results come through the providers' interfaces, which do not know a real customer's location the way the ChatGPT app on a phone does.

What the report covers

Four answers you did not have before.

Where you stand

How many AI answers name you and in what position, broken down by model. This is the starting value every later action can be measured against.

Who gets recommended instead

All named businesses with frequency and average position. This is your competition in AI answers, and it often differs from the list of competitors you have in your head.

What to do next

Prioritised actions based on your result: directories, reviews, structured data, external mentions, content that answers questions. In other words, the route from diagnosis to the next step.

Whether anything is moving

With a subscription we measure again automatically, weekly or daily, and changes arrive by email. So you can see whether actions work and when someone overtakes you.

Start free audit

Business name, industry, city and country are enough. No access to your website, no analytics connection.

Frequently asked questions
Isn't this just SEO?
Related, but not the same. Classic SEO works towards ranking positions in a list of results. Here the only thing that counts is whether a model names you in its answer. Different signals matter for that: consistent directory entries, reviews, mentions on trusted sites, structured data and content that answers a question directly.
Why do I see a different result when I ask ChatGPT myself?
Because AI answers are not deterministic: wording, model and timing all change the result. That is exactly why the audit measures across many prompts and several models instead of a single question. Your own one-question sample can turn out better or worse by chance.
What information do you need from me?
Business name, industry, city and country, optionally your website. For the free audit also an email address the report goes to. No access to your website, no analytics connection, no Google account.
How long does an audit take?
The free audit is usually finished in one to two minutes. Paid plans ask several dozen questions and take correspondingly longer, because each answer is evaluated individually.
Which languages and countries does it work for?
You choose the country and the language of the questions. At the moment the audit asks in German or English. If your customers ask in another language, the result shows your visibility for questions in English, not in that language.
Who is this worth it for?
Local businesses with high customer value, where the choice runs through a recommendation: law firms, medical and dental practices, tax advisers, trades, hotels. The more a single job is worth, the more expensive every enquiry that goes to a competitor without your knowledge.

Why ChatGPT, Gemini and Perplexity do not name your business

Detailed answers to the questions behind this: where AI answers get their names from, what decides the selection and what can be influenced.

Subscription: ongoing monitoring

Stay visible, month after month

AI answers change all the time. With ongoing monitoring you know straight away when a competitor overtakes you, and what to do about it.

plan/business

€49 /month

  • 25 buying-intent prompts
  • 4 AI models
  • Weekly monitoring
  • Competitor tracking
  • PDF reports
plan/proRecommended

€99 /month

  • All 30 buying-intent prompts
  • 6 AI models
  • Daily monitoring
  • Competitor tracking
  • PDF reports
  • Priority support

Billed by Paddle in euros · cancel monthly · VAT included in the EU, local tax may be added elsewhere