49 lines
3.2 KiB
Markdown
49 lines
3.2 KiB
Markdown
---
|
||
name: measure-engagement
|
||
description: Reviews community health metrics, sentiment, and experiment impact with recommended actions.
|
||
usage: /community-orchestration:measure-engagement --window 30d --detail full --dimensions "channel,persona"
|
||
---
|
||
|
||
# Command: measure-engagement
|
||
|
||
## Inputs
|
||
- **window** – time horizon (7d, 30d, 90d) for analysis.
|
||
- **detail** – summary | full report depth.
|
||
- **dimensions** – optional breakdown (channel, persona, program, cohort).
|
||
- **experiments** – optional list of programs to analyze.
|
||
- **alert_threshold** – optional metric threshold for escalations.
|
||
|
||
### GTM Agents Pattern & Plan Checklist
|
||
> Mirrors GTM Agents orchestrator blueprint @puerto/plugins/orchestrator/README.md#112-325.
|
||
|
||
- **Pattern selection**: Engagement measurement typically runs **pipeline** (data → diagnostics → sentiment → experiment readouts → actions). If diagnostics + experiment analysis can run in parallel, log a **diamond** segment with merge gate in the plan header.
|
||
- **Plan schema**: Save `.claude/plans/plan-<timestamp>.json` capturing window, data feeds, task IDs, dependency graph (analytics, CRM, sentiment tools), error handling, and success metrics (engagement %, advocacy, risk volume).
|
||
- **Tool hooks**: Reference `docs/gtm-essentials.md` stack—Serena for schema diffs, Context7 for platform documentation/conversation exports, Sequential Thinking for insights review cadence, Playwright for verifying dashboard/report embeds.
|
||
- **Guardrails**: Default retry limit = 2 for failed data pulls or sentiment processing; escalation ladder = Community Analyst → Community Lead → CS/Product leadership.
|
||
- **Review**: Run `docs/usage-guide.md#orchestration-best-practices-puerto-parity` before execution to confirm inputs, dependencies, deliverables.
|
||
|
||
## Workflow
|
||
1. **Data Pull** – aggregate platform analytics, CRM attribution, sentiment scores, and support signals.
|
||
2. **Health Diagnostics** – compute growth, activation, engagement, retention, and advocacy metrics by dimension.
|
||
3. **Sentiment Review** – scan community conversations, surveys, and NPS for emerging themes.
|
||
4. **Experiment Readouts** – evaluate running pilots against guardrails + KPIs.
|
||
5. **Action Recommendations** – produce prioritized playbook (content tweaks, ambassador outreach, escalations).
|
||
|
||
## Outputs
|
||
- Community health dashboard snapshot with annotated insights.
|
||
- Sentiment + risk report referencing `sentiment-analysis` findings.
|
||
- Recommended action list with owners, deadlines, and expected impact.
|
||
- Plan JSON entry stored/updated in `.claude/plans` for audit trail.
|
||
|
||
## Agent/Skill Invocations
|
||
- `community-analyst` – leads data aggregation and insights.
|
||
- `sentiment-analysis` skill – interprets qualitative signals and escalation needs.
|
||
- `community-lead` – receives action plan and drives follow-through.
|
||
|
||
## GTM Agents Safeguards
|
||
- **Fallback agents**: document substitutes (e.g., Community Lead covering Analyst) when analysts unavailable.
|
||
- **Escalation triggers**: if engagement drops below alert_threshold twice or sentiment exposes severe risks, escalate via GTM Agents rip-cord to cross-functional leadership.
|
||
- **Plan maintenance**: update plan JSON/change log when data sources, thresholds, or reporting cadences change to maintain GTM Agents audit parity.
|
||
|
||
---
|