home / skills / gtmagents / gtm-agents / attribution-playbook

attribution-playbook skill

safe

/plugins/marketing-analytics/skills/attribution-playbook

This skill guides attribution modeling and QA across campaigns, offering templates, governance, and change management to ensure accurate ROI measurement.

npx playbooks add skill gtmagents/gtm-agents --skill attribution-playbook

Review the files below or copy the command above to add this skill to your agents.

Files (1)

SKILL.md

1.1 KB

---
name: attribution-playbook
description: Standard operating guide for campaign tagging, attribution models, and
  QA.
---

# Attribution Playbook Skill

## When to Use
- Standing up or refreshing attribution models across teams.
- Training campaign owners on UTMs, tracking, and governance.
- Running QA before/after changing attribution weights or tooling.

## Framework
1. **Taxonomy Rules** – channel, campaign, creative, and experiment naming standards.
2. **Model Catalog** – when to use first/last/linear/position/data-driven with pros/cons.
3. **Data Contracts** – required fields, systems of record, sync cadence.
4. **QA Checklist** – sample records, dedupe logic, cross-tool reconciliation.
5. **Change Management** – approval workflow, communications, and measurement of impact.

## Templates
- UTM builder + validation sheet.
- Attribution model comparison matrix.
- QA runbook for campaign launches and quarterly audits.

## Tips
- Keep a shared "gotchas" log for recurring tagging errors.
- Align with finance on ROI/CAC definitions before rolling out model changes.
- Pair with `evaluate-attribution-models` for decision-ready narratives.

---

Overview

This skill is a production-ready attribution playbook that standardizes campaign tagging, attribution models, and QA for revenue teams. It consolidates taxonomy rules, a model catalog, data contracts, and operational runbooks to reduce leakage and speed decision-making. The goal is consistent, auditable attribution that aligns marketing, sales, and finance definitions.

How this skill works

The playbook defines naming standards and required fields, then maps use cases to recommended attribution models (first, last, linear, position, data-driven). It includes QA checklists, sample records, and reconciliation steps to validate pipelines across tools. Change management guidance and templates support safe model updates and stakeholder communication.

When to use it

Standing up or refreshing attribution across marketing and sales stacks
Training campaign owners on UTMs, tracking, and governance
Running QA before or after changing attribution weights or tools
Onboarding new analytics or ad platform integrations
Quarterly audits to reconcile cross-tool attribution and revenue data

Best practices

Enforce a shared taxonomy for channel, campaign, creative, and experiment names
Define data contracts with required fields, systems of record, and sync cadence
Keep a shared 'gotchas' log to capture recurring tagging and mapping errors
Align attribution changes with finance on ROI and CAC definitions before rollout
Use templates for UTM validation, model comparison, and QA runbooks to maintain repeatability

Example use cases

Choose an attribution model for a new multi-channel acquisition campaign using the model catalog and comparison matrix
Run pre-launch QA using the QA runbook to validate UTM fields, dedupe logic, and upstream ingestion
Reconcile paid media cost and conversion data across ad platforms and the CRM during a quarterly audit
Document and communicate an attribution model change using the change management workflow and approval checklist
Train campaign owners on UTM standards using the UTM builder and validation sheet

FAQ

How do I pick the right attribution model?

Use the model catalog: match simple awareness or short-funnel campaigns to first/last models and multi-touch or B2B pipelines to linear, position, or data-driven approaches; weigh pros and cons in the comparison matrix.

What essential fields should a data contract include?

Include canonical campaign id, channel, creative, timestamp, touch order, system of record, and sync cadence; specify dedupe rules and owner for each field.