How much of the conversation do you own?
Count how often AI assistants cite you against how often they cite competitors, then press Calculate to see what that gap could mean in revenue, using your own numbers. Runs entirely in your browser, no sign-up.
01 / How it works
Count your citations
Ask ChatGPT, Perplexity and Gemini the questions your buyers actually ask, and count how many answers mention or recommend you out of the total you tracked.
Enter your economics
Add your own estimate of how many buyers ask AI assistants each month, your average deal value, and your close rate. These stay fully editable.
Read the gap
Press Calculate to see your AI share of voice as a percentage, the split between buyers who hear about you versus competitors, and what your assumptions imply is at stake.
02 / FAQ
Questions, answered.
What does 'AI share of voice' mean?+
It's the share of buyer questions where an AI assistant mentions or recommends you, out of all the buyer questions you tracked. If you're cited in 3 of 20 questions, your AI share of voice is 15%. It says nothing about Google rankings, only about what ChatGPT, Perplexity and Gemini actually say back.
How do I actually measure how many times I'm cited?+
There's no shortcut: ask ChatGPT, Perplexity and Gemini the questions your buyers ask before they buy, and count how many answers mention you. Do this consistently, same questions, same assistants, and you have a real number to put into this calculator instead of a guess.
Why is this different from a search ranking?+
A search ranking is a position on a results page. AI share of voice is whether an assistant's generated answer names you at all. This tool makes no ranking claims and doesn't touch Google positions; it's strictly about how often AI answers cite you.
Is the revenue number a prediction?+
No. It's arithmetic on the inputs you typed: missed buyers times your close rate times your deal value. Change any input and press Calculate again to see the number update. Treat it as what your own assumptions imply, not a forecast of what will happen.
How many questions should I track for a fair sample?+
The default is 20. Fewer than that and one lucky or unlucky answer swings your percentage a lot; more gives a steadier read. Pull the questions from real buyer language: support tickets, sales call notes, comparison searches, not just guesses.
How do I actually improve my share of voice?+
AI assistants lean on what other sites say about you: reviews, comparison pages, structured content, mentions on sites they already trust. That off-site and on-site work is exactly what Beamcite runs for clients, aimed at getting cited more often in the answers that matter.
03 / More free tools
Turn 'us vs them' bullet points into a clean, semantic HTML comparison table AI engines can lift, plus matching JSON-LD.
Enter your domain and get a check-by-check verdict on whether your llms.txt actually follows the convention AI crawlers expect.
Paste a URL and see your full H1-H6 outline graded: one H1, no skipped levels, and how many headings are question-shaped.
Follow any URL's full redirect chain hop by hop and verify the final page's canonical tag points where it should.
Free strategy call · 30 minutes
Measuring is step one.
Beamcite does the ongoing work that moves this number: getting you cited more often by ChatGPT, Perplexity and Gemini, with a human running it.
Book the callNo payment. No obligation. A real human.