AI Brand Visibility Monitor

Track how visible your brand is when people ask ChatGPT, Claude, Gemini, or Perplexity for recommendations — in any language, not just English.

More and more customers are asking AI assistants "what's the best X" instead of googling it. If your brand doesn't come up in those answers, you're invisible to a growing share of buyers, and there's no dashboard for that yet. This Actor gives you one.

Who this is for

• Marketing / SEO teams running a GEO (Generative Engine Optimization) audit before or after a content push
• Founders who want a quick pulse-check: "does ChatGPT even know my product exists?"
• Agencies monitoring brand visibility for multiple clients across markets and languages
• Product teams tracking how a competitor launch shifts the AI-answer landscape

A practical scenario

Say you run a SaaS project-management tool and just published three "best PM tools" comparison articles. You want to know if that content actually moved the needle on how AI assistants describe your category.

1. Run this Actor with brandName: "YourTool", your top 3 competitors, and industry: "project management software".
2. Compare the visibility_score and average_position to a run from before your content push.
3. Re-run monthly to track the trend, not just a single snapshot.

What it measures

For each query (yours or auto-generated) x language x platform combination, the Actor asks the AI and analyzes the response for:

Metric	What it tells you
mention_count	How often your brand comes up
position	Where you rank when the answer is a list — the closest thing to a search-ranking equivalent for AI answers
sentiment	positive / neutral / negative tone toward your brand
share_of_voice	Your mentions vs. your named competitors' mentions, combined
cites_brand_domain	Whether your website is cited as a source (only computed if you provide brandDomain)
visibility_score (summary row)	A single 0-100 composite of all of the above

Why this Actor (vs. others in this space)

• Multi-language. Most AI-visibility tools only check English queries. If your customers ask in Turkish, Spanish, Arabic, or any other language, this Actor checks in that language too, not an English query with a translated brand name bolted on.
• No manual query writing required. Give us your industry and we auto-generate a solid first set of queries (best-tools, alternatives, recommendations, comparisons). Add your own exact queries any time with customQueries.
• Try before you pay. Set demoMode: true to see the exact output shape with zero API calls and zero cost.

Input reference

Field	Required	Type	Notes
brandName	Yes	string	The brand to track
brandAliases	No	array	Alternate names/abbreviations checked as the same brand
brandDomain	No	string	Enables citation detection (e.g. example.com)
industry	No	string	Powers auto-generated queries
competitors	No	array	Named brands to compute share of voice against
queryTemplates	No	array	best_tools, alternatives, recommendations, comparisons
customQueries	No	array	Your own exact questions, in addition to auto-generated ones
languages	No	array	e.g. ["en", "tr", "es"] - default ["en"]
platforms	No	array	openai, anthropic, google, perplexity
demoMode	No	boolean	Preview the output format, no real API calls, no cost

Output reference

Every query x platform check is one dataset row (see table above for field meaning), plus one summary row per run:

{
  "summary": true,
  "brand": "Nike",
  "total_checks": 8,
  "mention_rate_pct": 87.5,
  "average_position": 2.1,
  "positive_sentiment_pct": 75.0,
  "average_share_of_voice_pct": 24.0,
  "citation_rate_pct": 12.5,
  "visibility_score": 76.1
}

How it works

1. The Actor builds a query list: your customQueries plus anything generated from queryTemplates + industry + competitors.
2. Each query is sent to every platform you selected, in every language you selected.
3. The raw AI answer is passed back to the same AI platform for structured analysis (mention count, position, sentiment, competitor mentions, citation) - not a fragile regex, so it holds up across very different answer formats.
4. Results are written to the dataset one row at a time, plus a final summary row with the composite visibility_score.

Pricing

Pay per event, no subscription:

Event	Price
Brand query checked	$0.05
Actor start	$0.00005 (one-time, negligible)

Try demoMode: true first, free, to confirm the output shape before spending anything.

A note on accuracy

We don't invent numbers. Mention count, position, sentiment and citations come from asking the AI to analyze its own already-generated answer - nothing is guessed or hardcoded. If a response can't be parsed, that item is marked as no-mention rather than silently faked as a hit.

AI assistant answers are not deterministic - re-running the same query can produce a different answer even minutes apart. Treat a single run as a snapshot, and re-run periodically (weekly or monthly) to see a real trend rather than noise.

FAQ

Do I need my own OpenAI/Anthropic/Google/Perplexity API key? No. Pricing already includes the underlying AI cost - you only pay the per-query price above.

Can I check just one platform? Yes, set platforms to just the one(s) you want, e.g. ["google"].

What if my brand name is also a common word? Use brandAliases and, where possible, a more specific industry value to keep queries relevant and reduce false matches.

How is visibility_score calculated? A weighted blend: mention rate (30%), position (25%), positive sentiment (25%), share of voice (20%). It's meant for tracking your own trend over time, not as an absolute industry benchmark.