What this workflow produces: Two or more message variants testing a specific variable, with a prediction for which will perform better and why.
Choose one variable: subject line, opening hook, value prop angle, CTA style, or email length.
Run the A/B Message Testing prompt with your base message and variable specification. Output: 2-3 variants, each with a different angle on the specified variable.
Load into your sequencer with equal audience splits. Set minimum sample size of 30-50 sends per variant.
After 2 weeks: compare open and reply rates. Document which variant won and the insight for future messages.