A/B Conversion Rate Significance Calculator
Compare two independently assigned groups with a conversion significance test.
Measure conversions against eligible visitors or sessions under the counting rule you choose. Unique-person mode permits one conversion per person, while event mode allows repeated actions.
Measure conversions against an eligible traffic count using an explicit counting rule. The form distinguishes unique converted visitors from repeated conversion events, preventing the two rates from being interpreted as the same measure.
Decide which traffic can convert and what constitutes a qualifying outcome. Applying one rule to the numerator and another to the denominator can produce a misleading rate.
Use the same reporting period and attribution window for traffic and conversions. The calculator does not deduplicate users or link activity across devices.
Divide conversions by eligible traffic and multiply by 100. Unique-person mode gives the share who converted; event mode gives conversion events per 100 traffic units.
Sample inputs demonstrate the method. They are not a forecast or product offer.
The first example contains 50 unique converted visitors among 1,000 eligible visitors. Dividing 50 by 1,000 and multiplying by 100 gives a 5% visitor conversion rate.
Sample inputs demonstrate the method. They are not a forecast or product offer.
A rate above 100% is possible for repeated events because one visitor can generate several actions. It is not possible as a share of unique people who converted.
This descriptive rate does not determine whether differences between campaigns are statistically meaningful or caused by a particular change. Experiment design and uncertainty require separate analysis.
An event-per-visitor rate can when people trigger multiple events. A percentage of unique visitors who converted cannot exceed 100%.
Read our calculation and testing approach.