Guides
Measuring your brand's standing inside AI assistants
"Does ChatGPT know my brand?" is a fair question, but a yes-or-no answer says almost nothing. A brand may be cited one time in ten, always last, with a wrong detail; another may be recommended first on every question in its field. Between the two lies a measurable standing. Measuring it means replacing an impression with numbers you can track and compare.
Standing is not the same as a mention
Being mentioned once does not make a standing. What counts is regularity: across all the questions a customer in your field would ask, how often does your brand appear on its own, without its name being fed to the AI. A truly established brand is cited often, on varied questions, by several different AIs.
That spontaneous frequency is what separates a brand AIs really know from one they rebuild on demand. Type "tell me about [brand]" and an AI will almost always say something, even about a company it does not know — it makes things up. The real measure happens on questions where your name is not given.
The indicators that matter
A few simple indicators are enough to make standing objective. Presence rate: out of a hundred customer questions, how many bring up your brand. Position: are you named first, or buried at the end of the answer. Share of voice: against competitors, what fraction of the recommendations is yours. And consistency across AIs: are you known to all of them, or only one.
Taken together, these numbers give a visibility score far more telling than an isolated test. Above all they are comparable: from one quarter to the next, against a competitor, or before and after a communication effort, you can see whether standing is genuinely improving.
Why a number beats a feeling
Without measurement you rely on a few tries and intuition — often biased, because you test your own name and read the answers kindly. A figure computed over many questions and several AIs removes that subjectivity. It also lets you set a target and check that you are getting closer.
A brand that discovers it is cited on only fifteen percent of its field's questions knows exactly where it stands, and can track the effect of its actions month after month. That is the difference between "I think AIs know me" and "my presence went from fifteen to thirty percent in two quarters".
Getting a first measurement
You can approximate these numbers by hand: ask twenty to thirty industry questions to several AIs, count your appearances, note your average rank. It is instructive, but slow and hard to repeat identically for tracking over time.
Our free pre-diagnosis produces that first measurement automatically: a presence score on non-branded questions, in your market's language, with a glimpse of where your competitors stand. It is the quantified starting point from which tracking your standing becomes possible.
Frequently asked questions
How do I know my brand is really known to AIs, not just made up?
By asking questions where your name is not given. If the AI cites you on its own on your field's questions, it knows you. If it only talks about you when you type your name — sometimes with wrong details — that is reconstruction, not standing.
What counts as a good standing score?
There is no universal threshold: it depends on your field and the number of players. The value lies less in the absolute figure than in the comparison — against direct competitors and over time. A presence that rises quarter after quarter is the best signal.
Does this measurement replace classic reputation monitoring?
No, it complements it. Classic monitoring looks at what is said about you on the web and social media; the AI measurement looks at what assistants retain and recommend to your customers. The two are linked, but strong web presence does not guarantee strong AI presence.