A free tool for a clearer first picture
What does AI say about your practice?
Ask the questions your customers ask. Record whether your business is named and your website is cited. This free manual scorecard keeps those two measures separate, with a clear record of what you have - and haven’t - checked.
No sign-up. No AI calls. Your worksheet entries stay in this browser tab. Download them before you leave; refreshing clears the worksheet.
Enable JavaScript to use the calculator, or download the blank CSV worksheet and follow the guide below.
01 / Set your test
A few good questions. A repeatable check.
Start with two or three real customer questions. Include at least one that does not name your business. Choose the products and repeats before you start, then keep the same plan for your next comparison.
02 / Record what you can verify
Open each AI product yourself. Start a fresh conversation for each planned run, use the exact question, and save the full answer and visible sources. Select “Answer captured”, then assess the two checks independently. Leave anything uncertain as “Not assessed”.
All planned question types combined
These totals combine branded and discovery questions in your plan. Use the separate groups below to judge whether new customers might discover you.
- Business mention rate
- Not assessed
- Owned-site citation rate
- Not assessed
By question type and AI product
| Question group / product | Coverage | Business named | Owned website cited |
|---|
How to make this useful
- Test a real decision. Try a service and location question a new customer might ask. Keep questions that name your business in the branded group.
- Capture before you judge. Save the answer, source list, time, product and conditions. Count an explicit business name as a mention. Count a visible source link to your own domain as an owned-site citation. A third-party page about you is a separate kind of evidence.
- Use “No” carefully. Check the complete visible answer for a mention, and inspect its sources for your domain. A failed run or an uninspected source list is unknown.
- Keep the denominators. Each answer contributes at most once to each measure, regardless of how many times your name or link appears. Repeats are separate runs. Export your observations and blank plan to keep the method repeatable.
- Compare like with like. Repeat the same plan after a meaningful interval, recording website changes separately. If a model, mode, question or coverage changes, explain it before interpreting a change in rate.
What the numbers mean
Mention rate = assessed answers naming your business ÷ answers assessed for mentions. Owned-site citation rate = assessed answers linking to your website ÷ answers assessed for citations. The denominators can differ.
If four of eight planned answers have been checked for mentions, and one names you, the mention rate is 25% of four assessed answers. The other four are unknown. If only two source lists were assessed and one links to your website, the citation rate is 50% of two assessed answers.
This is a record of your manual observations, not an independently verified audit. Answers vary with time, product and conditions. Branded results do not prove discovery, and a mention or citation does not prove an enquiry or a sale.
Working in the CSV? Use captured, not-run or blocked for status; yes, no or a blank for each check. The browser export includes separate denominator and positive columns. Unknown positives are blank, not zero. The blank CSV is a worksheet, with no built-in formulas; it does not upload or reopen results in this page.
Google Search Console and Bing’s native AI reports measure different platform activity. Do not add their figures to this direct-product sample. Read our guide to AI visibility measurement for definitions, source guidance and comparison limits.
Worksheet v1 · Last reviewed 5 September 2026. Worksheet entries are held in memory only. Existing site analytics follow your cookie choice; the calculator does not send your entries to analytics. CSV files can include what you type, so keep downloads and supporting evidence private.
Want help interpreting the evidence?
Ask about a free visibility check
For a defined question set, captured evidence and a prioritised action plan, explore the £1,500 Visibility Audit.