💡 Platform Health Radar — Algorithm Shift Detection & Cross-Platform Content Pivot Alerts

[github-actions[bot]]

Summary

Monitor engagement patterns across all connected platforms to detect algorithm changes, declining organic reach, and platform instability — then alert the team with specific content pivot recommendations. Distinguishes "your content underperformed" from "the platform changed its algorithm" so lean teams don't waste time fixing the wrong problem. Transforms ContentTwin from a content tool into a social intelligence system.

Market Signal

TikTok's algorithm was retrained on U.S.-only data after the January 2026 Oracle-led restructuring — content moderation and creator prioritization shifted substantially (Teak Media 2026). Organic reach for text/image posts is declining across all platforms, pushing organizations toward video and new formats. Traditional monthly content planning is obsolete; "rough monthly themes with built-in flexibility for reactive content" is the new standard (Teak Media 2026). Enterprise social intelligence tools (Sprout's Trellis, Brandwatch, Meltwater) provide this capability for $500-5000+/mo — nothing exists at nonprofit pricing.

User Signal

Nonprofits posting content blind waste limited volunteer time on platforms where their reach has silently collapsed. The existing Editorial Calendar proposal (#377) plans WHAT to post but doesn't monitor WHETHER the plan is working or if platform conditions have changed. ContentTwin audit issues (#314-#326) show daily pipeline monitoring — the next step is monitoring whether the pipeline's output is actually reaching people.

Technical Opportunity

V1 uses published engagement metrics (likes, shares, comments, reach where available) tracked over time per platform. Statistical anomaly detection (rolling average deviation) identifies when engagement drops significantly. Cross-referencing drops across multiple orgs in the same cause area distinguishes algorithm changes (everyone dropped) from content problems (only you dropped). The system generates specific pivot recommendations: "Instagram reach dropped 35% this week across similar orgs — algorithm favoring Reels over static posts. Suggest: convert your next 3 planned image posts to Reel scripts using the Video Script Generator (#378)."

Assessment

Dimension	Score	Rationale
Feasibility	high	Uses published engagement metrics, not private APIs; anomaly detection is well-understood statistics
Impact	med	Saves wasted effort, but value scales with user base (cross-org aggregation is the real differentiator)
Urgency	high	TikTok algorithm just retrained; organic reach declining now; nonprofits are posting blind today

Adversarial Review

Strongest objection: This is just analytics, which Buffer and Hootsuite already offer on affordable plans. Anomaly detection requires significant data volume that a single nonprofit won't generate.

Rebuttal: Analytics tools tell you WHAT happened (engagement dropped 40%). The Radar tells you WHY — distinguishing content problems from platform algorithm shifts. The cross-org aggregation insight is the key differentiator: by anonymously comparing engagement patterns across ContentTwin users in similar cause areas, the system can detect platform-level changes that no single org's data would reveal. This is collective intelligence that individual analytics can never provide. Even without cross-org data, tracking your own engagement trends with automated alerts is a capability no affordable tool provides to nonprofits today.

Suggested Next Step

Define the engagement metrics schema per platform (Instagram: reach, impressions, engagement rate, Reel vs. static performance; Facebook: organic reach, engagement, video completion; LinkedIn: impressions, click-through; X: impressions, engagement rate; TikTok: views, completion rate, share rate). Design the anomaly detection algorithm (rolling 4-week average with 2-sigma threshold). Spec the alert format and pivot recommendation templates. Plan the anonymized cross-org aggregation architecture for v2.