AI Answer Monitor — Is Your Brand in AI Answers?

When your customers ask an AI ("best digital bank in Brazil", "top marketing agency for clinics"), does your brand come up? This Actor asks a real AI answer engine your questions and tells you, for each one, whether your brand is mentioned, how prominently, and whether your site is cited as a source. It's the search-visibility metric for the AI era — GEO / AI-SEO. No login.

Em português: monitor de GEO / AI-SEO, minha marca aparece no ChatGPT/Perplexity/Gemini, visibilidade de marca em respostas de IA.

Input example

{
  "brand": "Nubank",
  "domain": "nubank.com.br",
  "questions": [
    "melhor banco digital do Brasil",
    "qual conta digital sem tarifa vale a pena"
  ],
  "engine": "gemini"
}

• brand — the brand to look for in the answers.
• domain (optional) — your site, used to detect if the AI cites you as a source.
• questions — the questions your buyers would ask an AI (up to 100).
• engine — perplexity (cites sources) or gemini (Google Search grounding).

Output example

One row per question:

{
  "brand": "Nubank",
  "engine": "perplexity",
  "question": "melhor banco digital do Brasil",
  "checked": true,
  "mentioned": true,
  "mention_count": 4,
  "prominence": "top",
  "cited": false,
  "answer_excerpt": "Os mais citados no Brasil costumam ser Nubank, Inter e ...",
  "citations": []
}

• mentioned / mention_count — is the brand in the answer, and how many times.
• prominence — top / mid / low: where the first mention falls in the answer (top = named first = most visible).
• cited — the brand's domain appears among the answer's sources.
• answer_excerpt + citations — the raw evidence, so you can read exactly why.

What you learn

• Which questions you already win (brand named, top prominence) — protect and amplify these.
• Which questions you're invisible on (mentioned: false) — your GEO gap, where competitors are getting the AI's recommendation instead of you.
• Whether the AI trusts your site as a source (cited).

Run it weekly to watch your AI visibility move as you publish content.

Pricing (Pay per event)

Event	When	Price
actor-start	once per run	$0.01
question-checked	per question successfully asked to the AI	$0.10

Monitoring 20 questions = $0.01 + 20 × $0.10 = $2.01. A question whose engine is temporarily unavailable is returned with a note and not charged.

Engines

• Perplexity (sonar) — fast, web-grounded answers. Strong for the mention/prominence signal.
• Gemini (Google Search grounding) — answers grounded in Google Search, and returns the source links, so the cited signal is richest here. Also the closest proxy to what Google's AI surfaces to users.

mentioned and prominence work on both engines. cited depends on the engine returning source links (always on Gemini; Perplexity's base model may omit them).

Is this compliant?

Yes — the Actor sends your questions to a public AI API and analyzes the text of the answer. No scraping, no login, no private data.

Limitations

• AI answers are non-deterministic — the same question can vary run to run. For a trend, run the same set weekly and watch the direction, not a single snapshot.
• mentioned is a text match on the brand name — a very generic brand name can over- or under-count (use a distinctive name or brand form).
• Coverage and citation richness depend on the chosen engine (see above).

From the same maker

• B2B Leads Brazil · CNPJ Discovery — build and score the company lists behind your outreach.