A/B Test

Design → run → verdict, all on one page — the math is code, the AI narrates

  1. Listen
  2. Brief
  3. Variants
  4. A/B Test
  5. Budget
Design
power calc in code · α .05 · power .8

Guardian - Hàng Thật 100%

Metric
install rate (installs/clicks)
Sample / arm
12,779
Duration
~1 days
Variants in test
1
Run
◌ Simulated demo data
Arm A · control
8.40%
1,074 / 12,779 converted
Arm B · variant · winner
9.89%
1,264 / 12,779 converted
Readout
p = <0.001
SHIP ✓

We tested a 'Guardian - Hàng Thật 100%' badge/message against the original to see if it improved install rate from clicks. Variant B (with Guardian messaging) converted at 9.89% vs 8.40% for control, a relative lift of about 17.7%. With over 12,700 users per arm and a p-value effectively at 0 (z=4.12), this result is highly statistically significant, so we can be confident the improvement is real and not due to chance. Ship variant B (Guardian - Hàng Thật 100%) as the new default since it significantly and meaningfully improved install rate.

Reallocate budget →All tests