How Kolliq turns public creator data into decisions
Kolliq uses public YouTube observations to help creators understand what changed, what evidence supports the interpretation, what remains unknown, and what deserves testing next.
Creator-relative baselines
A video is compared with the creator's own history so a small channel and a large channel are not judged by the same raw threshold.
Age-adjusted performance
Newer uploads need time. Kolliq avoids overclassifying videos that have not had enough public observation time.
Competitor and cohort evidence
Competitors and peer cohorts help separate isolated channel movement from broader topic or market movement.
Confidence and unknowns
Kolliq labels insufficient samples, conflicting evidence, and private analytics that public data cannot verify.
Evidence first, not certainty theater
Public data helps narrow your next check. It cannot establish suppression, private CTR or retention changes, or an algorithm change without supporting evidence.
Definitions you can inspect
- Baseline
- A reference sample of comparable uploads. Record its size, format, and observation age. A median reduces the influence of one unusually large upload.
- Relative performance
- Observed views divided by the comparable baseline. An illustrative 30,000 views against 10,000 is 3x. A tiny baseline, format mismatch, or unequal age can make this ratio misleading.
- Confidence
- A description of evidence coverage and agreement, not a calibrated probability of success. Missing observations and contradictory cases reduce what a comparison can support.
- Outlier
- An unusually strong or weak result relative to the selected comparison. It is a candidate for investigation, not proof that a topic or format caused the result.
Methodology updates
6 September 2026: added explicit comparison definitions, worked examples, guide sources, and availability limits. These are explanatory clarifications, not a new validated prediction model.
Editorial and source standards