See how often AI names your brand, engine by engine.
Proofsource asks ChatGPT, Perplexity, Claude and Google AI Overviews the questions your buyers ask. It reports your mention rate, share of voice and average position on each engine, with a 95% confidence range on every number.
2 free scans, today and tomorrow. No card · no sales call
Mention rate is the share of buyer answers that name you.
Every scan puts the same buyer questions to every engine and keeps each answer. Mention rate is the answers that name your brand divided by all answers collected. In a public scan, AI named Tesla in 53 of 100 answers: a 53% mention rate.
- Counted on each engine and across all four
- Branded and unbranded questions are tagged, so you can read them apart
- Every count opens to the answers behind it
| Engine | Answers | Named Tesla | Rate | 95% range |
|---|---|---|---|---|
| ChatGPT | 25 | 16 | 64% | 45 to 80% |
| Google AI Overviews | 25 | 16 | 64% | 45 to 80% |
| Claude | 25 | 11 | 44% | 27 to 63% |
| Perplexity | 25 | 10 | 40% | 23 to 59% |
| All engines | 100 | 53 | 53% | 43 to 62% |
Every number carries a 95% Wilson interval.
25 answers is a small sample. ChatGPT named Tesla in 16 of 25, which is 64%, but the honest range is 45% to 80%. Proofsource prints that range beside the number, so a small dip does not read as a crisis and a real shift does not get waved away.
- Wilson score intervals stay sensible on small samples and near 0% or 100%
- Ranges narrow as answers accumulate
- The method is public, down to the formula
ChatGPT and Claude overlap between 45% and 63%. On 25 answers each, that gap is not yet proof that one engine favours Tesla more.
A leaderboard of the brands AI names instead of you.
Proofsource counts every brand the answers name, not only the ones you listed. In the Tesla scan, Rivian and SunPower each appeared in 13 answers, Ford in 10 and Enphase Energy in 9.
- Share of voice: your mentions divided by mentions of every tracked brand
- Average position: where in the answer you appear, 1 means named first
- Rivals found in answers are listed even if you never tracked them
| Brand | Answers naming | Share of voice | Tracked |
|---|---|---|---|
| Tesla | 53 | 54% | You |
| Rivian | 13 | 13% | Found in answers |
| SunPower | 13 | 13% | Found in answers |
| Ford | 10 | 10% | Found in answers |
| Enphase Energy | 9 | 9% | Found in answers |
Track it every day, split by the topics you sell into.
Custom plans ask every tracked question on every engine every day, so trends come from fresh answers, not one snapshot. Questions are grouped into topics, so you can see that you lead on one product line and trail on another. The free trial runs 2 free scans, today and tomorrow.
- Daily answers on Custom plans
- Topics drawn from what you sell
- Day-by-day trends per engine and per topic
| Day | Answers | Named | Rate |
|---|---|---|---|
| Mon | 100 | 52 | 52% |
| Tue | 100 | 55 | 55% |
| Wed | 100 | 49 | 49% |
Numbers you can take to a board meeting.
Ranges, not false precision
A 95% interval beside every rate tells you when a change is real and when it is sampling noise.
Four engines, four answers
In the Tesla scan, ChatGPT named the brand in 16 of 25 answers and Perplexity in 10. A blended score would hide that gap.
Rivals you did not list
Every brand AI recommends is counted, so the competitor you never thought to track still shows up.
Questions buyers ask
How does Proofsource measure AI visibility?
Proofsource asks ChatGPT, Perplexity, Claude and Google AI Overviews the questions your buyers ask and stores every answer. It counts which answers name your brand, where in the answer you appear and which pages are cited. From that it reports mention rate, share of voice and average position on each engine, each with a 95% Wilson interval.
What is share of voice in AI search?
Share of voice is your brand's mentions divided by the mentions of every tracked brand in the same answers. In a public Tesla scan, Tesla had 53 of the 98 mentions across five brands, a 54% share of voice.
What does average position mean?
Average position is where your brand appears in the answers that name it. Position 1 means the engine named you first. Lower is better, and it is only counted on answers that name you.
Why does Proofsource show a confidence interval?
AI answers vary from one asking to the next, and a scan samples a limited number of them. A 95% Wilson interval shows the range the true rate most likely sits in, so you can tell a real change from noise.
How often does Proofsource check AI answers?
Custom plans ask every tracked question on every engine every day. The free trial runs 2 free scans, today and tomorrow.
Which AI engines does Proofsource track?
ChatGPT, Perplexity, Claude and Google AI Overviews, on every plan. Grok and DeepSeek are available as add-ons on Custom plans.
Can I try Proofsource for free?
Yes. The free trial is 2 free scans, today and tomorrow: 25 buyer questions on ChatGPT, Perplexity, Claude and Google AI Overviews. No card. After the second scan, scanning pauses and your answers stay in your account.
How much does Proofsource cost?
Custom plans start from $64 per brand per month. Final price confirmed at launch. Every plan includes the four core engines; you choose how many questions and brands to track.
More of the platform
Prompts
The buyer questions you track, grouped by topic, tagged branded or unbranded, ranked by search demand, with the searches each engine runs.Explore Prompts
Citations
Every page AI cites in your category, the lists that shape answers, which ones mention you, and a stored copy of every answer.Explore Citations
Accuracy and sentiment
What AI says about you, how it frames you against rivals, which claims are wrong or out of date, and an audit of the profiles engines trust.Explore Accuracy and sentiment
Find out how often AI names you, today.
See whether AI names your brand, who it names instead, and what to fix. 2 free scans, today and tomorrow. No card · no sales call.
Last updated