# Ab Testing Framework

> Systematically test headlines, content formats, CTAs, channels, and posting timing for Bouts marketing with a one-variable-at-a-time discipline, minimum sample sizes, winner criteria, and immediate application of results. Use when running any Bouts content optimization test.

- Skill: `nickgallick/ab-testing-framework` (Agent Skill)
- Install (CLI): `npx skillmds add nickgallick/ab-testing-framework`
- Raw SKILL.md: https://api.skillmd.com/api/skills/nickgallick/ab-testing-framework/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Marketing & Growth
- Author: nickgallick (https://skillmd.com/u/nickgallick)
- Updated: 2026-09-08
- Page: https://skillmd.com/skills/nickgallick/ab-testing-framework

---


# A/B Testing Framework

## What to test

### Headlines
- Blog titles: data-driven vs question-based vs provocative
- Email subject lines: stat-heavy vs curiosity-driven vs urgent
- Tweet hooks: data point first vs question first vs statement first

### Formats
- Tweet threads vs single tweets for the same data
- Long-form LinkedIn vs short LinkedIn post
- Technical depth vs accessible summary for blog posts

### CTAs
- "Enter the Arena" vs "Try Bouts Free" vs "Benchmark Your Agent"
- "Get the Report" vs "See the Data" vs "Download the Index"

### Channels and timing
- Same content, different posting times
- Same content, different platforms
- Same content, different audience targeting

## Testing discipline
1. Test ONE variable at a time
2. Run for minimum 7 days or 1,000 impressions (whichever comes first)
3. Track: impressions, engagement rate, click-through rate, conversion rate
4. Winner criteria: >10% lift with consistent direction over the test period
5. Apply winner to ALL future content of that type immediately

## Test log format
| Test | Variable | Variant A | Variant B | Winner | Lift | Applied |
|------|----------|-----------|-----------|--------|------|---------|

## Current tests to prioritize at launch
1. Tweet hook: data point first vs question first (run in week 2)
2. Blog title: statement vs question (run in week 3)
3. Email subject: stat-heavy vs curiosity (run in week 4)