A brand people pick on a recommendation
Your customers ask who to choose, and more and more often they ask an assistant. You want to know what it tells them.
AI visibility.
Some people no longer ask Google, they ask an assistant. When someone asks ChatGPT, Gemini or Perplexity about a supplier in your field, either they mention you or they do not, and you usually have no idea which. AI visibility is the measurement of that state: which questions we tracked, how often you appeared in the answers, which pages the models drew on and who appeared instead of you. The output is a report and concrete recommendations, not a promise of placement. The initial analysis costs 300 €, regular monitoring starts at 80 € per month.
What it includesOr reach us directly: +421 902 074 830
Who it is for
Who AI visibility makes sense for
It makes sense where the customer chooses a supplier or a product and asks for a recommendation. It does not make sense where the decision comes down to the price on a comparison site or the branch around the corner.
A company whose organic traffic is falling
You hold your positions in Google, but the clicks keep dropping. AI Overview takes over part of the answer and you want to know whether it cites you or goes around you.
A B2B service with a long decision cycle
On a deal that takes weeks to decide, the assistant is often the first filter. Getting onto the shortlist is worth something different than a single click.
A brand after a redesign or a rebrand
You changed your name, your website or your focus and you want to check whether the models noticed and whether what they say about you is current.
What it includes
What the analysis covers
Measurement on specific questions from your field, not a general view of the brand.
01
The set of tracked questions
First we put together the questions a customer would really ask about your service, including the ones where they ask directly for a supplier recommendation. You approve the set, because the whole measurement depends on it.
02
Measurement across models
We ask the same questions in several assistants and record whether and how you appeared in the answer. A single measurement is a snapshot, so we also record the date and the wording of the answer.
03
The sources the models drew on
Where the answers name their sources, we find out which pages the model cited. It tends to be the most useful finding of the whole analysis, because it shows what kind of content gets you into the answers at all.
04
Comparison with competitors
Who appears instead of you and why. Whether it is a competitor, a directory, a comparison site or a third-party article, and how much of that you can influence.
05
Technical readiness of the website
Whether the website is reachable for AI crawlers at all: robots.txt and access for individual bots, structured data, consistency of your company details across the website and the llms.txt file. This is the part you fully control.
06
Recommendations and priorities
Concrete steps ordered by what has the best chance of changing something: what content to add, which pages to adjust and where your details are inconsistent. With no promise that they will put you first in the answers.
How we work
How it works
From building the questions to the report, and on to regular measurement if you want it.
-
01
Opening call
We go through what you sell, to whom, and who you want to be compared with. That gives us the first draft of the tracked questions.
-
02
Approving the set of questions
You review the set of questions and adjust it. The whole measurement rests on it, so we do not want to measure something that does not concern you.
-
03
Measurement
We ask the questions in several assistants and record the answers, the mentions and the cited sources, and we do the same for your competitors.
-
04
Report
You get a document with the current state, a comparison with your competitors, a list of the cited sources and recommendations ordered by priority.
-
05
Monitoring, if you want to continue
A single measurement is a snapshot. With monthly monitoring we repeat the same set of questions and track what changes, so that you see a trend and not one data point.
Principles
What we say plainly about this service
-
01
Placement in an answer cannot be bought or guaranteed
Nobody can promise you that ChatGPT will recommend you. What can be done is to measure the current state, find out which sources lead into the answers and remove the obstacles on your side.
-
02
Answers change and are not the same for everyone
The same prompt returns a different answer a week later, and it can return a different one to different people. That is why we record the date and the wording, and why monitoring says more than a one-off scan.
-
03
We measure, we do not produce a marketing number
The report contains what we actually measured, including the questions where you did not appear at all. If the result is unremarkable, we write that.
-
04
It is not a replacement for SEO
The models draw mostly on content that is on the web and that can be found. Good technical SEO and usable content are a precondition for this service, not a competitor to it.
Our stack
What we measure with
The answers of the assistants and data about what happens on your website. One without the other does not give the full picture.
Assistants
ChatGPT
Gemini
Perplexity
Google AI Overview
Website and data
Google Search Console
GA4
Crawler log analysis
Website readiness
robots.txt
Structured data
llms.txt
Collaboration
A one-off analysis, or tracking over time
one-off
Initial analysis
300 €
One complete measurement: the set of questions, measurement across assistants, the cited sources, a comparison with your competitors and a report with recommendations. A snapshot of the current state.
- Building and approving the set of questions
- Measurement in several assistants
- A list of the cited sources
- Comparison with competitors
- A check of the website's technical readiness
- A written report with priorities
Most common
basic package
Monitoring
from 80 € per month
It follows on from the initial analysis. We measure the same set of questions repeatedly, so that you see a trend and not one number. The size of the set and the frequency of measurement determine the final price.
- Repeated measurement of the same set of questions
- An overview of the changes against the previous period
- Competitor tracking
- A short monthly report
- Adjusting the set of questions as things develop
FAQ
Frequently asked questions.
What is AI visibility?
Brand visibility in the answers of AI assistants, meaning whether and how ChatGPT, Gemini, Perplexity or Google AI Overview mention you when someone asks them a question from your field. It is measured on a specific set of questions, not in general.
Can you guarantee that ChatGPT will recommend me?
No, and neither can anyone else. Placement in an answer cannot be bought. What we can do is measure the current state, find out which sources the models draw on for your questions and remove the obstacles on your website's side. The result is a probability, not a guarantee.
Why do you recommend monitoring instead of a single analysis?
The answers of the models change; the same prompt returns a different answer a month later. One measurement tells you how it looks today. Only repeating the same set of questions shows whether anything is improving and whether it relates to the changes you made.
How is this different from SEO?
SEO deals with positions and clicks in the search engine. AI visibility deals with mentions and citations in the answers of assistants, where a click often never happens. They overlap in that the models draw on content on the web, so SEO optimization is a precondition, not an alternative.
What exactly do I get in the initial analysis?
A written report with the set of tracked questions, the result of the measurement for each of them, a list of the sources the models cited, a comparison with your competitors, a check of the website's technical readiness and recommendations ordered by priority.
How much does it cost?
The initial analysis costs 300 € as a one-off. Monthly monitoring starts at 80 € in the basic package; the final price depends on the size of the tracked set of questions and the frequency of measurement. Monitoring follows on from the initial analysis. Prices for our other services are on the pricing page.
Who is it not worth it for?
For sales where the price on a comparison site decides, for purely local services where the customer chooses by distance, and for very narrow fields where the questions are practically never asked. We will say so after the opening call, before you order anything.
Do you need access to my website?
Not for measuring the answers themselves. For the technical readiness check and for the source analysis, access to Google Search Console and GA4 helps, and the server logs if they are available. The analysis can be done without them, it will just be less precise.
Is this related to the technical audit?
They partly overlap in the check of the website's technical readiness. A technical audit deals with the overall state of the website, AI visibility deals only with what affects mentions in AI answers. If you cannot decide, start with the audit.
Let's build something together
You bring the idea. We handle the rest.
Tell us what you need. We will reply within one business day with an honest take on whether we are the right fit, the scope and timeline we see, and a custom quote.