Knowledge Base
A practical help center covering the product, AI visibility concepts, tools, credits, reports, methodology, privacy, and troubleshooting. Use it before your first scan, after reading a report, or when you need to understand what a score or signal means.
Prompts & Fairness
Aitrack test prompts are informed by cross-platform market research and common real-world ways customers ask for recommendations. The aim is to represent genuine buying situations without shaping the test around one brand.
Market research across AI platforms shows recurring ways people discover, compare, and evaluate businesses. Common patterns include category discovery, location, use case, trust, price, service, occasion, and comparison.
These patterns are matched to the relevant business category and Singapore buying context. A pet store, ramen chain, dental clinic, tuition centre, and B2B software company therefore receive prompts that reflect different customer journeys.
A useful test prompt sounds like something a buyer might naturally ask an AI assistant before choosing a provider. It can include category, location, occasion, trust, price, or comparison criteria when those are common in the market.
The set is not written around the target company's exact selling points. It represents a balanced range of customer needs so a strong result reflects genuine visibility rather than favourable wording.
The same customer need can produce different answers across AI platforms. Comparing the prompt at a consistent level makes it easier to see where the brand appears, which competitors replace it, and what evidence is cited.
This prompt-level view helps Full Audit and Health Check connect visibility gaps to practical website actions.
The report shows the prompts tested and connects weak results to page-level fixes. A missed delivery prompt points toward delivery evidence; a missed trust prompt points toward reviews, credentials, case studies, policies, or FAQ content.
This turns market-informed prompts into practical evidence rather than a vanity score.