Your buyers are asking AI what to buy. Do you know what it answers?

SEO measures where your page sits on a results page. This measures whether an assistant names your brand when a buyer asks what to buy, which is a different question with a different answer. A page can rank first in Google and never be mentioned by ChatGPT, because the answer is generated from what the model has read about you across the web, not from your position in an index. A rank tracker cannot see any of it: there is no ranking to track, only an answer that either includes you or does not.

Then the report says so on the first line, and it is built for that case rather than tripped up by it: no empty charts, and every section that needs you in the answers steps aside for one that doesn't. You still get who gets named in your category instead of you, what those brands are praised for, the questions that went to a competitor, and the sites those answers are built from. Including whether your own domain is one of them, which is the difference between nobody recommending you and nobody reading you.

That is the argument for measuring now rather than later. The brands assistants name today are the ones they have accumulated the most material about, and accumulating it takes time, so this is a position built over quarters rather than bought in a week. A baseline now tells you where you stand while the answers are still moving and while most of your category has not looked, and it gives you something real to compare against when they do.

Yes, and it matters more than it sounds. You pick the market you sell in, and we write and ask the buyer questions in that market's language: Danish questions for Denmark, German for Germany. Assistants recommend a noticeably different set of brands depending on the language they are asked in, because they draw on different reviews, comparisons and forum threads. A report run in English would be measuring a market you don't sell in.

Every report ends with a prioritized action plan: what to fix first, ranked by how much of your visibility gap each item accounts for, with the specific number from your run behind it. Not generic advice. Findings like which buyer question type you are absent from, which attributes competitors get described with and you never do, and which sites the assistants read when answering questions in your category.

You can, and you'll learn something from it — one question, asked once, is one data point. AI answers vary between platforms and between runs of the same question, so a single check can't tell you whether you were unlucky or genuinely absent, and it can't track a competitor, break results down by buyer intent, or give you a number to compare against next quarter. The report is 50 to 200 real buyer questions, asked of four assistants, scored and aggregated with a stated margin of error — the difference between a spot check and a measurement.

Every report states its own margin of error. The score comes from a sample of AI answers, so it carries sampling error like any measurement. We show the 95% confidence range alongside the number, so if you re-measure later you can tell whether a change is real or just noise. Bigger plans ask more questions and therefore produce a tighter range.

Answers shift as models are updated and as more content about you and your competitors appears online. Most people start with one report as a baseline, then re-check quarterly. Re-running for the same brand adds a trend comparison against your previous score, so you see the direction and not just the level.

No. We use your brand, category, and competitor names only to generate your report. Nothing you submit is used to train any AI model.

ChatGPT, Claude, Gemini, and Perplexity. If a platform fails to answer during your run, the report says so and excludes it rather than quietly blending a partial result into your score.

A few minutes, depending on plan size: roughly two minutes for Starter and up to about ten for Pro, since every question is asked of all four assistants in turn. You don't need to wait on the page. We email it the moment it's ready.