Methodology

Visual Radar scores are comparative signals, not final judgments of creative quality. The system favors recent public activity, peer-relative performance, data completeness and integrity checks over one raw number such as follower count.

VIS is a creator-level signal. It combines reach, engagement, viral power, consistency and momentum, then applies activity and integrity factors. The result is designed to identify creators whose public work is both visible and repeatedly active.

Rising focuses on recent growth. A creator needs at least one qualified recent work, enough profile completeness and enough metric history across time before the score is considered eligible. The model weighs relative and absolute follower growth, reach acceleration, breakout works and publishing regularity.

Viral ranks individual visual works. It compares reach against follower scale, measures engagement quality from public interactions, considers velocity and absolute reach, and limits eligibility to recent works so older posts do not dominate the current view.

When possible, comparisons happen inside category and follower-tier cohorts. If a cohort is too small, the fallback widens to category-level or global-tier comparisons so the ranking remains stable instead of pretending there is precise peer data.

• Private metrics and private messages are not used
• X and Apify availability can delay or fail collection
• Incomplete public metrics lower confidence rather than being guessed
• Scores may change when a new valid snapshot is collected