Originality Score
After render, before upload: a 0–100 score summarises the video's risk. Six independent signals are combined into a single reading, so you decide with information — the score does not block publishing.
Why it matters
YouTube doesn't publish the internal score it uses to decide demonetization — but channels that fall share common patterns. Originality Score turns those patterns into auditable metrics. You see the risk before YouTube does.
What gets measured
The score combines signals about the script, the rendered video, and the channel's history. Each signal contributes a fixed weight — no single one sets the band, the reading is aggregated.
- AI detectionThe script is analysed by specialised external detectors. The more human it sounds, the higher the signal.
- Structural varianceComparison of this script with the channel's recent videos. Measures narrative-structure repetition, not just words.
- Editorial decisionsCount of completed editorial checkpoints. Videos with active curation score higher.
- Lexical diversityVocabulary richness of the script. Penalises overly repetitive scripts.
- Asset repetitionReuse of images/clips versus the channel's recent videos.
- Pacing varianceCut pacing compared to the channel average. Avoids mono-pacing and extreme outliers.
What happens with the result
The score lands in one of four bands: safe, warning, high risk, and critical risk. In every band the scorer points to the signals pulling it down and suggests specific fixes — the decision to publish stays yours. No score guarantees compliance: YouTube's policy is judged by them, not by us.
Plan availability
Every plan computes the score and sees the breakdown. Overriding via written justification requires Studio or Scale — Creator must regenerate or edit.