The GroVirally Video Intelligence Framework
GVIF turns a video into 39 individually scored factors, rolled up into 8 weighted dimensions and one overall score. Every weight below is fixed and versioned — nothing is decided by the AI at analysis time.
GroVirally does not ask AI for a random viral score.
An AI that's simply asked 'will this go viral, 0-100?' produces a plausible-sounding but unreproducible guess — ask it twice and you can get two different numbers, with no way to know why. GVIF never does this. The AI's only job is to describe what it observed; a fixed, versioned scoring engine — code, not a model — is the only thing that ever computes a number.
Video
Your upload or supported URL
Objective Measurements
Duration, cuts, aspect ratio, audio levels
AI Interpretation
Gemini describes what it observed, per factor
GVIF Framework
Fixed, versioned dimension & factor weights
Deterministic Scoring Engine
Our math — never the AI — computes the score
Video Intelligence Report
Scores, findings, and recommendations
Every dimension, in depth
Click a dimension to see its purpose, what's measured, example signals, weak vs. strong implementations, a sample recommendation, and exactly how its score is computed.
How the overall score reads
- Exceptional90+
- Strong80–89
- Promising70–79
- Needs Improvement60–69
- Major Revision Recommended40–59
- Weak Creative Foundation0–39
These bands describe structural readiness, not a virality guarantee — see AI Transparency for how confidence is calculated.