playbooks

home / skills / gtmagents / gtm-agents / nurture-testing

nurture-testing skill

/plugins/lead-nurture-orchestration/skills/nurture-testing

This skill guides planning, executing, and logging nurture experiments and regression tests to ensure quality, compliance, and measurable outcomes.

npx playbooks add skill gtmagents/gtm-agents --skill nurture-testing

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

Files (1)
SKILL.md
1.2 KB
---
name: nurture-testing
description: Use when planning, executing, and logging nurture experiments and regression
  tests.
---

# Nurture Testing & QA Skill

## When to Use
- Running experiments (subject lines, timing, offers, personalization) in nurture flows.
- Performing regression tests before go-live or updates.
- Documenting QA evidence for compliance/audits.

## Framework
1. **Test Design** – hypothesis, variable, control, variant, KPI, sample size, duration.
2. **QA Checklist** – links, tokens, segmentation, tracking, device/browser coverage, accessibility.
3. **Evidence Logging** – screenshots, seed inbox captures, MAP logs, CRM task verification.
4. **Significance & Rollout** – evaluation method (frequentist/Bayesian), rollout criteria, holdback rules.
5. **Regression Cadence** – schedule for periodic audits, triggered when assets/tokens change.

## Templates
- Experiment brief + tracker.
- QA matrix (scenario, expected result, status, owner).
- Evidence archive folder structure.

## Tips
- Always include control groups or holdouts for long-running nurtures.
- Automate reminders to re-run QA after major MAP/CRM updates.
- Pair with `personalization-logic` and `marketing-ops-partner` to catch edge cases early.

---

Overview

This skill helps teams plan, execute, and document nurture experiments and regression tests for marketing automation and CRM-driven flows. It standardizes test design, QA checks, evidence capture, and rollout rules so experiments run reliably and audits have clear proof. Use it to reduce risk during updates and to validate personalization and segmentation at scale.

How this skill works

The skill guides you through a five-step framework: test design (hypothesis, variables, KPIs, sample sizing), a QA checklist (links, tokens, segmentation, tracking, accessibility), evidence logging (screenshots, seed inbox captures, MAP/CRM logs), significance evaluation (frequentist or Bayesian) and controlled rollout rules. It also defines a regression cadence for recurring audits and triggers when assets or tokens change. Templates for briefs, QA matrices, and evidence archives make execution repeatable and auditable.

When to use it

  • Planning A/B or multivariate experiments in nurture flows (subject lines, timing, offers, personalization).
  • Running pre-launch regression tests after creative, MAP, or CRM changes.
  • Documenting QA evidence for compliance, audits, or stakeholder review.
  • Checking segmentation, tokens, and tracking across device/browser combinations.
  • Scheduling periodic audits for long-running or evergreen nurtures.

Best practices

  • Define a clear hypothesis, control, variant, KPI, sample size and test duration before execution.
  • Always include a control or holdout group for long-running nurtures to measure baseline performance.
  • Use the QA checklist to validate links, dynamic tokens, tracking pixels, and segmentation logic.
  • Capture and store evidence (screenshots, seed inboxes, MAP logs, CRM tasks) in a structured archive.
  • Automate reminders to re-run QA after major MAP/CRM updates and when assets or tokens change.

Example use cases

  • Experiment brief and tracker to compare two subject line strategies across segments.
  • Pre-release regression run to validate dynamic personalization tokens and fallback content.
  • Evidence package for an audit showing seed inbox captures, tracking logs, and QA matrix.
  • Routine cadence to re-audit nurture flows after platform upgrades or template changes.
  • Pairing with personalization logic and marketing ops partners to catch edge cases early.

FAQ

What evidence should I collect for audits?

Collect screenshots, seeded inbox captures, MAP send logs, tracking pixel confirmations, and CRM task or lead records. Store them in a structured evidence archive.

How do I decide between frequentist and Bayesian evaluation?

Choose frequentist methods for traditional hypothesis tests and fixed-horizon experiments; use Bayesian approaches for adaptive experiments or when you want continuous probability updates during rollout.