Manager guidesArticle
How to read the visibility measurement in CitationLab AI Marketing Manager
What do mentions, citations, Share of Visibility and position mean? How to read your first numbers from the AI Monitor module, and what you should not overreact to.
Published approx. 4 min read
In this article+–
Curious how AI talks about your brand?Run a free visibility check
The first numbers from the AI Monitor module are ready. There is a lot to look at, and some of the numbers behave differently from what you are used to in Google. This guide explains what the numbers mean, which ones to look at first, and what you should not draw conclusions from.
The short version: First, look at how often you are mentioned and cited compared with your competitors. Then look at the differences between models and between stages of the buying journey. Judge the development across several measurements, since single measurements fluctuate.
Mentions and citations are two different things
The most important distinction to understand is between mentions and citations:
- Mention: the AI model names your brand in the answer itself. This is where the customer reads your name.
- Citation: the AI model points to a page as a source. Citations can point to your website, or to other pages that talk about you, such as comparison sites, articles and directories.
You can be mentioned without being cited, and cited without being mentioned. Best is both: the AI recommends you and also points to your website as a source.
The numbers to look at first
Visibility index
A single number from 0 to 100 that combines how often you are cited and how often you are mentioned. Use it as a quick temperature check and to follow the development over time.
Share of Visibility
Your share of visibility compared with the competitors you track. This is your market share in AI answers, and the number that best shows whether you are gaining or losing ground. Manager shows both Citation Share, your share of the citations, and Mention Share, your share of the mentions.
Visibility Score
An overall assessment of your position, divided into four levels: Dominant, Competitive, Present and Weak. The level gives you a quick answer to where you stand in the market.
Avg Position
How high your sources rank in the list of sources the AI model shows. 1 is best. A high position means the model gives more weight to your content.
Sentiment
How you are talked about, on a scale from negative to positive. You find the details in the Sentiment tab. Negative mentions are worth reading in full, because they often point to a source you can do something about.
Look at the differences, not just the average
The overall number often hides the most interesting part. Use the tabs in the AI Monitor module to break it down:
- Models: ChatGPT, Gemini, Perplexity and Google answer differently and use different sources. It is normal to be visible in one model and almost invisible in another.
- Prompt Matrix: visibility per business area and per stage of the buying journey. Many businesses are visible in early questions and drop out when the customer asks who to choose.
- Citations: which domains and pages are cited, and whether your own are among them.
- Fan-Out Map: which sub-questions the AI model breaks the question into before it answers. It shows what your content needs to answer to be used.
What you should not overreact to
AI answers are generated from scratch every time, and the sources the models use change constantly. One measurement can therefore show something completely different from the next without anything having really changed. Three pieces of advice:
- Look at the trend across several measurements. Manager tests whether a change is real. If it is not, it says within noise, and you should hold off on a conclusion.
- Compare with your competitors. If everyone drops at the same time, the model has probably changed. If only you drop, it is worth looking into.
- Low visibility is a finding. Many businesses are almost invisible in AI search in the first measurement. That is the starting point you improve from, and the reason it is worth measuring.
Three questions you should be able to answer after the first measurement
- How often are we mentioned when the customer asks without using our name?
- Who is mentioned instead of us, and in which questions?
- Which sources do the AI models use when they answer in our area?
We cover the second and third questions in the next guide: Who AI recommends instead of you, and why.
Next guide
Frequently asked questions
Why are my numbers low?
Most businesses start low, because AI models pull information from different sources than the ones the business has worked on for Google. Low numbers at the start are common and give you a clear starting point.
How often are the numbers updated?
The numbers update every time the prompts run, at the frequency you have chosen. Weekly measurement gives a good picture of how things develop.
Why don't the numbers match what I see when I ask ChatGPT myself?
The answers vary each time the question is asked, and they are affected by settings and chat history. Manager measures systematically across many runs, which gives a more reliable picture than single questions.
What is the difference between Share of Visibility and Share of Model?
The terms are used for the same thing: your share of the visibility in AI answers compared with your competitors. In Manager it is called Share of Visibility.
Was this page helpful?