skills.rest
ericosiu
ericosiuCOMMUNITY
Explore all skills by this author →
GitHub

Growth Engine

Create A/B tests, score variant results with bootstrap confidence intervals and Mann-Whitney U tests.

3.3k|656|Updated Mar 28, 2026
Search Tags:#statistical analysis#ab testing#bootstrap ci#mann-whitney u#growth automation#marketing experimentation#playbook optimization
One-click install
npx skills add https://github.com/ericosiu/ai-marketing-skills --skill growth-engine-ericosiu
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: Growth Engine
Source: https://github.com/ericosiu/ai-marketing-skills/tree/main/growth-engine
Command: npx skills add https://github.com/ericosiu/ai-marketing-skills --skill growth-engine-ericosiu

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires numpy, scipy, urllib.

What problem does it solve?

This Skill solves the problem of marketing teams guessing what works by turning campaigns into a repeatable experimentation loop that measures results and promotes statistically supported winners.

Core Features & Use Cases

  • Experiment Engine (A/B and batch): Create experiments with hypotheses, track variants, and keep running until each variant reaches a minimum sample threshold.
  • Statistical scoring + confidence: Score outcomes using bootstrap confidence intervals and a Mann-Whitney U test to decide between keep, trending, and discard/crash.
  • Auto-playbook + next tests: Promote winning variants into a living playbook and suggest what you should test next based on what’s already been evaluated.
  • Weekly scorecards + pacing alerts: Generate a cross-channel weekly report and monitor campaign health (pipeline staging and recruiting velocity), with optional API integrations.

Quick Start

Use the Growth Engine to create, log, and score a content experiment by telling an AI agent to run a two-variant test for the metric you care about, then log metrics after publishing and finally score to update the playbook.

Frequently Asked Questions about Growth Engine

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate A/B testing for marketing experiments across multiple channels?▼

Marketing experimentation automation lets you create A/B or batch tests with hypotheses, track variant performance, and collect data until reaching a minimum sample threshold. It applies to growth loops across content, email, SEO, and recruiting channels.

What statistical methods are used to score A/B test results for marketing campaigns?▼

Statistical scoring for A/B test results uses bootstrap confidence intervals and a Mann-Whitney U test to evaluate variant performance. This combination decides whether to keep, trend, or discard variants based on configurable experiment thresholds.

How do I set up a YAML-driven experiment threshold for growth automation?▼

Growth automation uses a YAML-driven skill entry point to configure experiment thresholds for creating, logging, and scoring tests. You define minimum sample sizes and variant parameters to automate the marketing experimentation loop.

Can I generate weekly scorecards and pacing alerts for cross-channel campaign reporting?▼

Weekly scorecards and pacing alerts monitor cross-channel campaign health by generating reports on pipeline staging and recruiting velocity. Optional API integrations allow automated tracking of campaign performance and progress.

How do I promote winning A/B test variants into a living playbook?▼

Winning A/B test variants are promoted into a living playbook after statistical scoring confirms their performance. The playbook then suggests what you should test next based on previously evaluated variants and outcomes.

Does marketing experimentation work with numpy and scipy for statistical analysis?▼

Marketing experimentation relies on numpy, scipy, and urllib as dependencies for statistical analysis and data collection. These libraries enable bootstrap confidence intervals and Mann-Whitney U testing to score variant performance accurately.

Related Skills

Explore More Marketing & Sales Skills→
MakisuoMakisuo
1.6k

clickhousectl-local-dev

Install and configure a local ClickHouse environment with clickhousectl.

Community
Intermediate
LazyIsEfficientLazyIsEfficient
15

growth-engine

Automates growth experimentation with A/B and multivariate tests, scoring experiments and generating weekly scorecards.

Official
Advanced
8GG-Git8GG-Git

recruiting-pipeline

Track candidates through recruiting pipeline stages from sourcing to acceptance.

Community
Intermediate
OpenLaborOpenLabor
4

Growth Experiment Engine

Identify and track A/B experiments with bootstrap confidence intervals and Mann-Whitney U tests.

Official
Advanced