In short
- No list of names
- There is no ranking here. It ages within a quarter and says nothing about your market and your questions.
- The sharpest question
- Does every finding come with an address, or with a recommendation? That difference is the whole difference.
- The cheapest test
- Ask for the same measurement to be run twice. If the outcome differs without explanation, there is nothing to steer on.
Which seven questions expose an offer?
The right-hand column is not a caricature but what quotes really say. Each of those sentences can be true and none of them can be checked — and that is the point: you are buying a measurement, and a measurement you cannot verify is an opinion with a chart.
| Question | A usable answer | An answer that makes you leave |
|---|---|---|
| Which questions do you ask? | a list I see beforehand and may adjust | “our AI generates relevant prompts” |
| Which engines do you measure on? | a list with names and what happens when one drops out | “all major AI platforms” |
| How do you weigh between engines? | per market, with source and month | no weighting, or one that is not explained |
| What do I get with a finding? | the address of the source that caused it | “build authority in your industry” |
| Can I repeat the measurement? | yes, with the same set, and the difference is stated | a new question set every month |
| What do you not measure? | a list, named unprompted | “we cover AI visibility completely” |
| Do you also execute? | a clear boundary: what yes, what no, and who signs | “fully automatic, you have to do nothing” |
Why is there no ranked list of names here?
For three reasons, and the first is the most honest: we sell such a service ourselves, so a ranking from us is not an independent judgement. The second is that such a list is out of date within a quarter and then becomes a page someone believes. The third is that naming something you have not measured is exactly the mistake this guide warns against.
Can I do this without a tool?
For the first measurement, certainly. Write down ten questions your customers ask, leave your own name out, and put them to three assistants. Note per question whether you were named and which sources sat underneath. Half a day of work, and you know more than a dashboard tells you in a month.
A tool becomes useful once repetition is the work: the same set, every month, across several markets, with the difference beside it. Keeping that up by hand works twice and then stops.
What if I want to switch later?
Ask up front whether you can take your question list and your history with you. It looks like a detail and it is the only reason switching is expensive: without your old measurements you start over at zero and lose the one thing that made them worth having, the comparison with yourself.
Frequently asked questions
Why do you not name competitors?
Because naming something we have not measured is exactly what we hold against others. Where we do say something about another provider, a date and a source come with it and it concerns something publicly checkable — their own files, their own pricing page. An opinion about a product we have not used does not belong in a guide.
What if a provider answers all seven well?
Then the choice is about something else, usually about whether they also do the work beyond the measurement. That is a separate decision with its own price and its own risk — another guide covers that.
Read on
Free, no account and no card.