---
name: social-copy-ab-testing
description: Design controlled social copy experiments and publish or schedule test variants through BulkPublish with measurable hypotheses.
---

# Social Copy A/B Testing

Create comparable copy variants that test one meaningful variable at a time.

## Workflow

- Define the hypothesis, audience, platform, primary metric, test window, sample or posting plan, and stop condition.
- Vary one element such as hook, CTA, proof point, framing, or format; keep the rest as consistent as practical.
- Check character limits and platform requirements with `platform-reference`.
- Use separate BulkPublish posts or platform-specific content only when the comparison remains interpretable; label variants clearly.
- Use `get-analytics` after the test window and report supported metrics, sample limitations, and the winning interpretation without overstating causality.

Create drafts first unless the user explicitly authorizes scheduling or publishing.