Discover the questions your customers ask AI, and who it recommends.
Track real questions on ChatGPT, Perplexity, Gemini and Claude. See whether you are mentioned, which competitors are named instead, and the websites each answer is built from.
From question to opportunity, step by step.
Find the questions
Get ideas from your industry, locations, competitors and the questions already on your website, or let AI suggest what customers ask before they buy.
Ask every engine
Each question goes to ChatGPT, Perplexity, Gemini and Claude with web search on, for the location you choose. Tracked questions are asked again automatically.
See who gets named
Read the full answer with your brand and competitors highlighted, the order they appear in, and every source the engine cited.
Find the gaps
Questions where competitors are recommended and you are not, ranked by a transparent opportunity score.
Fix them
Send gaps to the AI Visibility Fixer, write better pages in the Content Optimizer, and get onto the cited sites with the Citation Builder.
Measure the change
Your AI Recommendation Rate in the Fixer comes from these answers, so you can see whether your fixes actually moved AI.
No guessed scores. Ever.
Engines without a connection are shown as "Not tracked". Mention rates appear only after enough answers, with the count next to them. Every number links back to the answers it came from.
Prompt Research questions
Where do the answers come from?
From each AI company's official API with web search switched on: the OpenAI Responses API with web search for ChatGPT, the Perplexity Sonar API, the Gemini API with Grounding with Google Search, and the Claude API with its web search tool. We store the full answer text and the sources each engine cites.
Is that exactly what my customers see?
Not always. The consumer apps can personalise answers by location, history and account, and models change over time. API answers with web search are a consistent, repeatable way to measure, so treat them as a strong signal, not a screenshot of every person's screen. We label them as API answers everywhere.
How is the mention rate calculated?
It is the share of stored answers that name your business or cite your website. We only show it after at least 3 answers, and always show how many answers it is based on. With fewer answers we say so instead of guessing.
What is the opportunity score?
Intent weight × how often competitors appear × how often you don't. A "ready to buy" or local question where competitors are named and you never are scores highest. The formula is shown in the dashboard next to the number.
Which locations and languages work?
Any. Add a city or country to a question and the engines are asked for that place. Questions can be in any language your customers use.
What does each plan include?
Free: up to 5 prompts and a few manual test answers a month. Starter: 10 auto-tracked prompts checked monthly. Professional: 25 prompts checked weekly. Agency: 80 prompts checked weekly, shared across client projects. Exact limits are on the pricing page.
Find out who AI recommends for your customers' questions.
Free plan available. Any city, any country, any language.