People do not just Google things any more. They ask an AI, and we measure whether the answer mentions you at all.
Search is shifting quietly, without a headline. More and more decisions now start in a chat window rather than on a results page, and what comes back is one recommendation instead of ten links. If the models have never heard of your domain, you did not lose that comparison. You were never in it. Your SEO tools cannot see this, because it has nothing to do with your site. It has to do with what a model learned long before it ever met you. So we built a number that shows you where you stand.
From a domain name to a clear score in under a minute. Here is what happens in between.
Just the domain. No setup, no code snippet, no access to your site. It works the same for your own site, a client's site, or a competitor you have never met.
Claude and GPT get eight questions each about your domain, and no permission to look anything up. A model that can search will find any live site in a second and tell you it knows you, which proves nothing. We want the part that is already baked in.
Each model is asked three times and we keep the middle answer, so one lucky or unlucky run cannot drag your result around. The two models are scored separately, then combined into one number from 0 to 100. You always see both, not just the blend.
Add the domain to your dashboard and we re-scan it every week, whether you are looking or not. Expect a fairly flat line, and know that this is the honest result. A model's knowledge does not change between Tuesdays. The line is there for the day a new model version arrives, because that is the moment a score genuinely moves, and it is the one worth catching.
One number, three zones. Here is how to read your result.
The models do not reliably recognize the domain. When someone asks AI for a solution in your field, you are simply not in the answer. There is no quick fix for this, and we will not pretend otherwise. But it is worth knowing that a growing channel has never heard of you.
The models know you exist and roughly what you do, but the details are thin or out of date. This is also the riskiest zone. A model that half knows you is the one most likely to state something confidently wrong about you.
The models describe you accurately and bring you up on their own, without being asked about you by name. Still worth watching, since every new model version is trained from scratch and can shift this in either direction.
One scan takes about thirty seconds. Scanning is free for any domain, always, with no account and no card.
Scan My Domain