How VisoTrak audits AI visibility.

An answer slot is a place where an AI platform gives a business recommendation for a customer-style prompt. VisoTrak tests which business appears in that slot, whether it is your business, and whether the answer gives the customer a clear path to act.

For each industry, we run a curated set of recommendation prompts that mirror real customer language — best, emergency, affordable, premium, and specialty queries. For every prompt, we check three things:

• Whether the business is mentioned at all.
• Where it ranks relative to competitors.
• Whether the AI surfaced a clear next action — website, phone, or booking path.

These five platforms cover the substantial majority of where customers ask AI for local recommendations today. Each has a distinct retrieval, ranking, and grounding model — so being recommended on one is no guarantee of being recommended on the others.

• ChatGPTOpenAI

  Conversational, broad coverage. Frequently used for shortlist queries.

• PerplexityPerplexity AI

  Search-grounded. Strong actionability with citations.

• ClaudeAnthropic

  Long-context reasoning. Used for comparison-style prompts.

• GeminiGoogle

  Tied to Google index. Bridges classic local SEO and AI search.

• GrokxAI

  Real-time signals from X. Used for trending recommendations.

We maintain per-industry prompt sets reflecting how customers actually phrase recommendation queries — best, emergency, affordable, premium, and specialty variants — and we localize them by city. A dentist in Houston gets a different prompt set than a lawyer in Austin, because customers ask differently in each category. We expand and refine prompt sets as new patterns emerge in AI usage.

The overall AI Score is the average across all prompt × model checks. Tier thresholds are Critical <40, At Risk 40–59, Visible 60–84, Dominant 85+.

A business can be mentioned by an AI without being recommended in a way a customer can act on. If the AI does not surface a website, phone number, or booking path, the customer has nowhere to go — so the mention does not convert. We score these mentions as amber, because the visibility exists but the action does not.

Each AI platform answers customer questions in a different way. ChatGPT and Claude draw mostly on knowledge from their training data, so they reward businesses that have a clear, consistent presence across the web. Perplexity and Grok search the live web at query time, so they reward businesses that show up in current articles, directories, and reviews. Gemini blends both.

That means the same business can rank #1 on one platform and not appear at all on another. This is signal, not noise. A business that is strong on training-grounded platforms but weak on live-grounded ones may have strong historical authority but stale current coverage. The reverse can mean recent activity has not yet shaped how trained models see you.

Your overall score weighs each platform equally. As more platforms come online, your score reflects a broader truth — sometimes higher, sometimes lower. Both directions are honest.

That is why one platform can rank you #1 while another misses you completely. VisoTrak shows the disagreement so you know where visibility is strong, weak, or stale. The phrasing below describes what each platform tends to reward — not its internal algorithm, which is proprietary and changes.

• ChatGPTOpenAI
  Training-grounded

  Tends to reward clear, consistent authority across the web — businesses with a stable, well-described presence in widely-indexed sources.

• PerplexityPerplexity AI
  Citation-grounded

  Tends to reward current citations, live web evidence, and fresh sources — businesses that appear in recent articles, reviews, and directories.

• ClaudeAnthropic
  Trust-grounded

  Tends to reward structured, trustworthy explanations and clear business context — businesses with services, pricing, and policies plainly stated.

• GeminiGoogle
  Local / entity-grounded

  Tends to reward local and entity signals, Google-connected context, and service clarity — businesses with strong place / category coverage.

• GrokxAI
  Real-time-grounded

  Tends to reward real-time web signals, recency, and social or contextual momentum — businesses being talked about right now.

Phrased as tendencies, not internal mechanics

AI platforms are non-deterministic and continuously updated. Models change, indexes refresh, and competitor content evolves. A score is a snapshot, not a permanent rating. We recommend re-running the audit monthly and tracking movement with a Pulse Report.

VisoTrak is not a sales funnel for any single agency or website builder. Our scoring rubric is the same for every business, every industry. We surface available solution paths after the audit — including do-it-yourself, freelancer, and full-service options — and we will continue to add others as the market matures.