AI Visibility Check — Is ChatGPT Recommending Your Business?

When locals ask ChatGPT or Perplexity for "the best gym near Cincinnati," your business either comes up or a competitor gets the customer. This Actor asks both assistants live, with web search on, exactly like a real searcher — and returns what actually happened: how many answers named you, who was recommended instead, and which local sites the AI pulled from. One check costs $0.79 and takes about 10–30 seconds.

No invented metrics. You get real observed counts — "appeared in 4 of 6 answers, ChatGPT 3/3, Perplexity 1/3" — because that's what the assistants actually said, and that's the only honest way to measure this.

What you get back

{
  "businessName": "Graeter's Ice Cream",
  "location": "Cincinnati, OH",
  "asked": 6, "sampled": 6, "appeared": 2,
  "engines": [
    { "engine": "ChatGPT", "sampled": 3, "appeared": 1 },
    { "engine": "Perplexity", "sampled": 3, "appeared": 1 }
  ],
  "recommendedInstead": ["Aglamesis Bro's Ice Cream", "Hello Honey"],
  "citedDomains": ["cincinnatirefined.com", "365cincinnati.com", "yelp.com"],
  "label": "Asked live with web search near Cincinnati"
}

• appeared / sampled — real appearance counts over live category questions, split per engine.
• recommendedInstead — the businesses the AI named instead of you. This is who is winning your customers in AI search.
• citedDomains — the local sites the AI answers drew from. Getting listed there is the concrete lever to change the answer.

Who uses this

• Business owners: find out in one run whether AI search knows you exist.
• Marketing agencies: run it on a prospect before the pitch — "ChatGPT recommended your competitor in 3 of 3 answers" opens doors.
• Local SEO consultants: citedDomains is a ready-made citations worklist, sourced from what the AI actually reads.
• AI agents: callable via the Apify MCP server; returns structured JSON and finishes inside a normal tool-call window.

Input

{
  "businessName": "Graeter's Ice Cream",
  "location": "Cincinnati, OH",
  "businessType": "ice cream shop",
  "website": "https://www.graeters.com"
}

businessType is required — the assistants are asked category questions, so without it there is nothing to check. website is recommended: a cited domain counts as an appearance even when the answer words your name differently. Spell the business name the way Google knows it.

Pricing

One visibility-check event ($0.79) per delivered check. Failed runs are never charged — if no AI answer could be sampled, you pay nothing.

Not what this does

• It does not produce a 0–100 "AI SEO" rating — no such number exists in reality, and we won't fabricate one.
• It does not track rankings over time in one run — each run is a fresh live sample (schedule it weekly if you want a time series).
• It does not fix your visibility — it tells you who's winning and which sites the AI reads, which is where the fixing starts.
• Not for national brands with no local market: the questions are category-plus-city, so a business without a local footprint isn't a fit.

For AI agents (MCP)

Agents can run this check directly: add Apify's MCP server as a connector — https://mcp.apify.com/?actors=moonlings/ai-visibility-check — and it appears as the moonlings--ai-visibility-check tool, metered through your Apify account like any other run. Useful as a scheduled watchdog: have your agent re-check weekly and flag when the appearance counts move.

FAQ

Why do counts differ between runs? AI answers genuinely vary — that's the point of sampling live. Each run reports exactly what was said this time.

Which engines? ChatGPT and Perplexity, both with live web search, asked as a searcher in the business's own city.

Want the full picture? The AI Visibility Report adds the competitor, review, and ads evidence behind the answer, up to a full graded research report. The Review Gap Check measures the review signals the assistants lean on. The Thumbtack Scraper pulls the local-pro competitive field — ratings, hires, and reviews by service and city. The AI Fact-Check goes further — it checks whether what AI says about a business (hours, address, phone) is actually correct.

Who this is for

Local business owners, marketing agencies, and local SEO / AEO (answer engine optimization) practitioners who need AI search visibility ground truth: brand monitoring across ChatGPT and Perplexity for "near me" buyer queries, citation sources worth targeting (GEO), and honest appearance counts to put in a client report. A lightweight alternative to domain-level AI brand visibility trackers when the question is a local market, not a global brand.

Works with AI agents

Agent-callable end to end: run it from any agent runtime via the Apify API or MCP, and it is x402-payable — an autonomous agent can discover, pay for, and run a check with no account. Structured JSON out, ready to pipe onward.