A/B Testing Sequences

Before enabling A/B:

• Primary sequence is created
• Variant B sequence exists
• Both sequences have at least one step
• Variant B is active when executing

In Sequence Settings:

1. Enable A/B test.
2. Select Variant B Sequence.
3. Set Split Percentage (1-99%) for Variant B.

Example: 30% means 30% of enrolled contacts go to B and 70% stay in A.

During API/internal execution enrollment, each contact is assigned:

• Variant A (primary sequence)
• Variant B (configured variant sequence)

Assignment is random using your split percentage.

• If Variant B is inactive or has no steps, enrollment is blocked.
• Keep variant names explicit (for example, Post-Visit vA and Post-Visit vB).
• Change only one major variable at a time (timing, message style, channel order).

Track differences in:

• Completion rate
• Review conversion
• Detractor resolution speed
• Response quality

Use Reports Overview and sequence metadata to compare outcomes over a meaningful sample size.