4,084of 4,246 rows in scope
Experiment

Variant comparison

Every arm of the split test side by side. Click a variant to focus the drill-down panels below on that group.

Performance ranking

Conversion rate with 95% Wilson interval

Variant B16.23% · 325/2,002
Variant A13.88% · 289/2,082

Session duration distribution by variant

Box plot — median, IQR and 1.5×IQR whiskers

Variant Amedian 1m 19s · IQR 48s2m 11s · n=2,082
Variant Bmedian 1m 31s · IQR 56s2m 24s · n=2,002

Full comparison matrix

All computed variant metrics under the active filters

VariantSessionsTraffic shareConversionsConv. rate95% CIAvg. durationMedianAbs. liftRel. lift
Variant A2,08251.0%28913.88%12.46%15.43%2m 18s1m 19s
Variant B2,00249.0%32516.23%14.68%17.91%2m 35s1m 31s+2.35%+17.0%
Drill-down

All variants combined

4,084 sessions · 15.03% conversion

Device breakdown

Sessions and conversion rate within the focused group

  • desktop16.12%
  • mobile13.34%
  • tablet17.07%
  • unknown11.32%

Traffic source breakdown

Sessions and conversion rate within the focused group

  • organic13.80%
  • direct15.23%
  • paid_search16.99%
  • social13.28%
  • email17.76%
  • referral13.86%
Lift summary

Variant B shows an absolute difference of +2.35% pts and a relative lift of +17.0% against Variant A. The 95% confidence interval for the difference spans +0.16% to +4.55% percentage points.