Buyers ask AI first now. We show you if it says your name.
Search used to be ten blue links you could rank for and measure. Today people ask an AI for “the best tool for X” and get one answer. If you're not in it, you don't exist for that buyer, and almost nobody can see it happening.
The problem
Analytics tools see Google. They don't see the answer a buyer got from an AI five minutes ago, so brands optimise blind.
What we do
We ask Claude and Gemini the questions your buyers ask, read every answer, and turn it into a score, a list of who wins instead of you, and a plan.
Who it's for
Founders, creators, and small teams who sell something and want to know, in plain language, whether AI recommends them.
How the score is calculated
Your 0–100 score blends four things we can measure in every answer. There are no hidden multipliers: this is the whole formula.
For example, being named in half of all answers, always in the top three, and never cited or praised scores about 40. Being named in every answer, always first, with your site cited and praised, scores 100.
- Mention rate60%
How often an AI engine names your brand when asked a buyer question. This matters most.
- Top-3 placement20%
How often you're in the first three recommendations, because people act on the top of the list.
- Your site cited12%
How often the engine links to your own website as a source.
- Positive tone8%
How often the way you're described is clearly positive rather than neutral or negative.
What we hold ourselves to
We show our work
Every score comes with the real answers behind it, so you can read what the AI actually said instead of trusting a number.
We don't flatter you
Questions never include your brand name. Naming you in the prompt would make the score look better than reality, so we block it.
We say when we're unsure
If an engine answered from its own memory instead of live search, we label it. Scores are estimates, and we call them that.
We steady the noise
AI answers change run to run. Paid scans ask each question several times and average the results, and we show how often repeats agreed.
Nobody can buy a better score
Results can't be paid for or tuned by us. If your score goes up, it's because the answers changed.
We don't ask for your code
Citeably only needs your public website. No repository access, no passwords, nothing private.
What Citeably can't tell you
- • Exactly what any one person saw. Answers differ by person, place, and moment. We measure the pattern, not one screen.
- • That a fix will work. We don't control the engines, so we point you to what tends to matter and then re-measure to show you what changed.
- • Anything about engines we don't query. Our reports only cover Claude and Gemini.
See where you stand in a minute.
Type your website. We read it, ask the engines, and show you the real answers. Free, no signup.