Visibility

Visibility

Answer the only question that matters in AI search: when a buyer asks an AI who to choose, does your brand come up — or your competitor?

Inputs

• Brand name + domain (ask if not given).
• 3-5 "money questions" a buyer would ask (derive from the audit report's top queries if present, else generate from the site's services and confirm with the user).
• DataForSEO MCP connected.

Steps

1. Build the question set: 5 buyer-intent questions, e.g. "best <category> in <city>", "who should I hire for <service>", "<product> vs alternatives". Conversational phrasing, the way people talk to a chatbot.
2. Check Google AI Overviews (DataForSEO SERP API): for each question, fetch the Google SERP and inspect the AI Overview block. Record: does an AI Overview appear, is the brand cited in it, which competitors are cited.
3. Check LLM answers (DataForSEO LLM/AI endpoints where available on your plan): query each question against the available AI answer endpoints and record brand mentions vs competitor mentions. Where an endpoint is not on the plan, run the question as a standard organic SERP and use top-10 presence as the proxy, and SAY it is a proxy.
4. Check the brand query: search the exact brand name. Is the site ranking #1 for itself, does a knowledge panel exist, do third-party pages (reviews, directories, LinkedIn) corroborate the brand? AI systems lean on corroboration — a brand that only exists on its own site gets skipped.
5. Score it /100:
   • 40 pts: brand cited in AI answers/Overviews for money questions (8 per question)
   • 25 pts: competitor citation gap (full points if you appear at least as often as the top competitor)
   • 20 pts: corroboration (third-party mentions, reviews, structured citations)
   • 15 pts: machine readability (llms.txt, schema, crawlable, AI bots not blocked — pull from the audit report)
6. Explain the skips: for every question where the brand was absent, name who showed up instead and the most likely reason (no answering content, no corroboration, blocked crawler, thin schema).

Output

AI VISIBILITY: <brand> — <date>
SCORE: <n>/100

PER QUESTION
"<question>" — AI Overview: <cited/skipped> · LLM answers: <cited/skipped> · who shows instead: <names>

WHY YOU'RE SKIPPED
1. <reason> — <evidence>

SCORE BREAKDOWN: citations <n>/40 · competitor gap <n>/25 · corroboration <n>/20 · readability <n>/15

Save to seo-reports/visibility-<date>.md. The diagnose skill consumes it.

Rules

• Real responses only. If an endpoint fails or is unavailable, mark that question "NOT CHECKED" — never simulate an AI answer.
• Always name the competitors that ARE being recommended. That line is what makes the report actionable (and shareable).