mcpField
P
performance-metrics

performance-metrics

github.com/gtmagentsMarketing & growth43067 installsUpdated 30w ago
Repository

About performance-metrics

A marketing skill for defining and tracking social performance metrics, intended for campaign owners and stakeholder reporting. It covers metric frameworks, attribution tags, alerts, reporting cadence, and experiment measurement.

Key features

  1. 01Build awareness, engagement, and conversion metrics
  2. 02Standardize UTM and content tags
  3. 03Set alert thresholds and severity
  4. 04Plan reporting cadence and formats
  5. 05Record experiment KPIs and decision rules

Use cases

  1. 01Set KPI targets for new campaigns
  2. 02Build dashboards or scorecards
  3. 03Diagnose reach or conversion issues mid-campaign
  4. 04Summarize experiments and define next steps

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/gtmagents/gtm-agents --skill performance-metrics

After installing, type/performance-metricsin Claude Code to confirm it loaded.

Details

From the project README

Social Performance Metrics Skill

When to Use

  • Setting KPI targets for new campaigns or evergreen programs.
  • Building dashboards or scorecards for exec/stakeholder updates.
  • Diagnosing pacing, reach, or conversion issues mid-flight.

Framework

  1. Metric Stack – map awareness (reach, impressions), engagement (ER, saves, shares), conversion (traffic, pipeline), and community health (sentiment, response time).
  2. Attribution & Tagging – enforce UTM structure, post IDs, and partner tags for downstream analysis.
  3. Alerting Rules – define thresholds (± variance vs goal), alert channels, and severity tiers.
  4. Reporting Cadence – daily pacing snapshots, weekly recaps, campaign retros, quarterly trendlines.
  5. Experiment Measurement – document test KPIs, guardrails, and success decision trees.

Templates

  • KPI scorecard (metric → target → actual → variance → owner → notes).
  • Alert playbook (trigger condition → notification → response steps).
  • Retro template summarizing learnings, insights, and backlog actions.

Tips

  • Normalize metrics per channel and audience to avoid false comparisons.
  • Pair with analytics warehouse data for revenue-attribution context.
  • Keep metadata consistent (pillar, campaign, creative ID) so insights are reusable.

Comments

Markdown supported