Transparency
How the Meritum Score works
A score nobody can inspect is a score nobody should trust. Every weight in the model is published here.
Model MERITUM-SCORE-V1.0
Content Quality
20%
Depth, originality and durability of published work across platforms.
Engagement Quality
20%
Meaningful responses and discussion, not raw impression counts.
Consistency
15%
Sustained contribution rhythm over time rather than isolated spikes.
Audience Growth
10%
Organic audience development. Deliberately a minor factor.
Community Contribution
15%
Replies, support and participation in other creators' work.
Cross-Platform Impact
10%
Contribution recognized across several connected networks.
Authenticity
10%
Signal integrity as assessed by the quality engine.
Design principles
- Contribution over popularity. Follower count is deliberately a minor input; audience growth is capped at 10% of the model.
- Determinism. The same inputs always produce the same score. No hidden randomness, no model drift between page loads.
- Honest gaps. Platforms expose different data. Where a metric is unavailable, Meritum labels it as unavailable rather than estimating it.
- Attributability. Every point traces back to a dimension, a weight and an underlying set of contribution events.
Known limitations of this proof of concept
- Demo mode generates stable synthetic data when platform credentials are absent.
- The quality engine is heuristic and is not production-grade fraud detection.
- No MTI is minted, transferred or claimable; reward figures are illustrative.
- Scores are not comparable to any live Meritum deployment.