track-source — Claude Code command for campaign attribution
track-source combines partner campaign data and produces attribution reporting with revenue contributions and suggested changes. Use it to review co-marketing results.
Apache-2.0 ★ 450 updated 2026-04-03
.claude/commands/track-source.mdWhat track-source does
The command moves through data collection, normalization, attribution, insight generation, and reporting. It reconciles partner CRM and analytics records, then applies the chosen model to allocate pipeline or revenue across partners and channels. Its deliverables include a dashboard snapshot, a contribution table against targets, and an action list tied to assets or channels.
Provide a campaign identifier, analysis window, and report depth. Optional inputs let you split results by partner, channel, asset, or region, or override the attribution model. The command stores a plan and change history under .claude/plans. Both partners’ data should use shared taxonomy and UTM parameters; otherwise, attribution across sources can be contested.
What it helps with
- reconcile partner CRM and analytics records
- allocate pipeline or revenue across channels
- identify strong assets and audience segments
- prepare a campaign dashboard and executive summary
- keep an auditable plan of reporting changes
Example prompts for the agent
- Run a full attribution report for the “AI Launch Tour” campaign over 30 days with
/partner-co-marketing-orchestration:track-source --campaign "AI Launch Tour" --window 30d --detail full. - Compare partner and channel contributions for the “AI Launch Tour” campaign over 7 days with
/partner-co-marketing-orchestration:track-source --campaign "AI Launch Tour" --window 7d --detail full. - Create a summary for the full campaign period with
/partner-co-marketing-orchestration:track-source --campaign "AI Launch Tour" --window campaign duration --detail summary.
How to install track-source
Manually: put the files in place
.claude/commands/track-source.mdCall it with /track-source. For all projects use ~/.claude/commands/
Manually: put the files in place
~/.codex/prompts/track-source.mdCall it with /prompts:track-source. Codex marks this format deprecated in favour of skills
Manually: put the files in place
.github/prompts/track-source.prompt.mdManually: put the files in place
.gemini/commands/track-source.tomlGemini stores commands as TOML: the prompt text goes into the prompt field
Questions about track-source
How do I install track-source in Claude Code?
Place the command file at .claude/commands/track-source.md in your project. Invoke it with /partner-co-marketing-orchestration:track-source.
What does track-source do?
It combines partner CRM and analytics data, applies an attribution model, and produces a dashboard, contribution table, and optimization actions.
What data does track-source need?
It works with analytics and CRM data from both partners, using shared taxonomy and UTM parameters. You can also set the reporting window, detail level, breakdowns, and attribution model.
Source text
track-source.md · Apache-2.0
--- name: track-source description: Aligns attribution, dashboards, and ROI reporting for partner co-marketing campaigns. usage: /partner-co-marketing-orchestration:track-source --campaign "AI Launch Tour" --window 30d --detail full --- # Command: track-source ## Inputs - **campaign** – joint campaign identifier matching briefs/assets. - **window** – time range for analysis (7d, 30d, campaign duration). - **detail** – summary | full report depth. - **dimensions** – optional breakdowns (partner, channel, asset, region). - **attribution_model** – optional override (first-touch, last-touch, multi-touch). ### GTM Agents Pattern & Plan Checklist > Mirrors GTM Agents orchestrator blueprint @puerto/plugins/orchestrator/README.md#112-325. - **Pattern selection**: Tracking/reporting usually runs **pipeline** (gather → normalize → attribute → insights → reporting). If normalization + attribution can run in parallel (separate partners), log a **diamond** segment with merge gate. - **Plan schema**: Save `.claude/plans/plan-<timestamp>.json` with campaign, data sources, task IDs, dependency graph (CRM, MAP, analytics), error handling, and success metrics (attribution confidence, SLA adherence, partner satisfaction). - **Tool hooks**: Reference `docs/gtm-essentials.md` stack—Serena for data model diffs, Context7 for partner taxonomy docs, Sequential Thinking for readout cadence, Playwright for dashboard QA if embedded. - **Guardrails**: Default retry limit = 2 for failed data pulls or reconciliation mismatches; escalation ladder = Co-marketing Analytics Lead → RevOps/Finance partners → Exec sponsors. - **Review**: Run `docs/usage-guide.md#orchestration-best-practices-puerto-parity` before execution to confirm agents, dependencies, deliverables. ## Workflow 1. **Data Gathering** – ingest analytics/CRM data from both partners using shared taxonomy + UTM parameters. 2. **Normalization** – reconcile schemas, currencies, and funnel stages; dedupe overlapping records. 3. **Attribution Calculation** – apply agreed model to allocate pipeline/revenue across partners + channels. 4. **Insight Generation** – highlight top-performing assets, audience segments, and drop-off points. 5. **Reporting Package** – produce dashboards, executive summary, and recommended optimizations. ## Outputs - Attribution dashboard snapshot with partner/source breakdown. - Revenue/pipeline contribution table vs targets. - Optimization action list referencing asset or channel adjustments. - Plan JSON entry stored/updated in `.claude/plans` for audit trail. ## Agent/Skill Invocations - `co-marketing-analytics-lead` – validates data + insights. - `attribution` skill – enforces methodology consistency. - `partner-marketer` – receives recommendations for next sprint. ## GTM Agents Safeguards - **Fallback agents**: note substitutes (e.g., Partner Marketer covering Analytics Lead) if specialists unavailable. - **Escalation triggers**: if attribution accuracy drops below guardrails twice or partners dispute sourced pipeline, escalate to revenue/partner governance board per GTM Agents rip-cord. - **Plan maintenance**: update plan JSON/change log when models, data sources, or reporting cadences change to maintain audit parity. ---
© gtmagents · Apache-2.0 · gtmagents/gtm-agents