Every package, price and measurement count on one page.

They do, and we are open about that. One question to one AI at one moment tells you very little. That is why we put eight to forty different buying questions to six platforms at once: 48 to 240 separate measurements. Every question asked and every result measured is in the report, so you can check it and repeat it yourself. What you get is a baseline at a fixed moment with a method that can be repeated, not a guarantee of a position. That honesty is exactly what makes it useful: measure again in six weeks with the same questions and the difference comes from your changes.

Go ahead, but you will be measuring something else. Your own ChatGPT knows your history and your location, and often already knows who you are. We ask blind: your brand name and domain do not appear in the question, so you see what a buyer sees who does not know you yet. And one answer is one platform at one moment. With Semrush, six engines named the brand 203 times in 240 answers, but Perplexity least often and Claude most. You only see that spread if you measure all six.

Three things you cannot find out on your own. First, whether AI names you and how often, broken down per platform. Second, which competitors get named in your place, with the reason the AI gave itself. Third, the sources those answers were built on: which sites and pages the AI used. That leads to a list of actions ordered by expected impact, where each action says what to do, where, and how you can tell it worked.

We cannot prove that for your market, and we are not going to claim it. What we can show: when somebody does ask, they get one answer with two or three names in it, and the AI has already done the comparing. There is no page two. In our Semrush measurement, 240 answers produced exactly two recommendations while the brand was named in 203. Whoever is in those two wins the business. The question is not whether it is big yet, but whether you are in there when it is.

The measurement runs at the moment you order. Nothing is pulled from a database: your questions go to the six platforms right then, and the answers are minutes old by the time the report reaches you. That is why the measurement date is on the cover. Because AI models keep refreshing their sources, repeating after a few weeks is worthwhile; that is what the Monitor is for, with twelve monthly measurements at €19 per report.

That is when the report earns its keep. Every action describes the handling, the page it applies to, and a check you can run yourself to confirm it worked. Not loose terms like 'implement schema', but what needs to be where and how to verify it. Actions are ranked by expected impact and effort required, so you start at the top. We measured our own site and scored zero; that report is on this site, including the list we started working through ourselves.