Contingency Engine
Automated engagement scoring and propagation logic · Sample data, not live
Engagement contingency engine
📝
Content posted
→
📊
First-24h engagement scored
👆 clicks · 📩 sends · ⭐ ratings
→
⚖️
Score vs. threshold
>75 = HIGH
50–75 = MEDIUM
<50 = LOW
→
HIGH → Propagate to all matching + 2 adjacent channels
MEDIUM → Keep on origin channels, re-test next cycle
LOW → Deprioritize, AI rewrites the angle, queue for review
→
🔁
Cycle repeats hourly
What this means in practice
Border alert (Tuesday, score 92): auto-propagated from Sahel-AR to Saba-AR + Anwar Health-AR. +1,840 new reach. No human action needed.
Service spotlight (Sunday, score 41): held on origin channels only. AI suggests rewriting the headline — "Verified service" vs "New service" performs +18% in A/B history (see TEST-05).
Survey #12 wave 2 (today): high-tier targeting unlocked because quality rating from wave 1 = 4.4 / 5, drop-off = 19% (below 30% threshold). System advanced to medium-tier broadcast automatically.
Algorithm controls
Auto-propagate ON Auto-deprioritize ON AI rewrite suggestions ON Quality threshold for auto-propagation: ⭐ 3.8 / 5