Most marketing tools call any A/B test with a higher number on one side a 'winner'. We do the math properly. Tests run until they hit either a Bayesian confidence threshold you set or a minimum sample size — whichever comes first — and the platform tells you when a result is not statistically meaningful instead of pretending it is.
Each capability ships on the standard plan. Nothing in this list is an upsell or a separate product line.
Some teams want a p-value, others want the probability that variant B beats variant A. We compute both, show the credible interval, and warn explicitly when sample sizes are too small for either method to mean anything.
Test subject, hero image, and CTA simultaneously with a fractional-factorial design that uses far fewer recipients than a naive full grid. Results show main effects and interactions separately.
Hold back 5% of recipients and measure what would have happened without any campaign. Indispensable for showing leadership that the cost of the sequence is justified by incremental revenue, not background conversion.
Run the same test on email and SMS to see which channel each variant performs best on — sometimes the 'losing' subject in email is the winning SMS opener.
Winning variants are promoted automatically into the parent flow or drip; losing variants archive with their full result attached so a year from now you don't re-run a test that already failed.
Trigger multi-step journeys from product events without writing glue code in your application.
Scheduled multi-message campaigns with time-zone awareness and behavioral exits.
Build segments from product events and user attributes; targets stay live as data changes.
Reach customers on the channel they actually open — email, US-native SMS, WhatsApp Business, and any system you can point a webhook at.
Know what happened to every message — and why — without piecing it together from carrier exports and inbox screenshots.
