Estimating required duration and traffic
This is a calculator that answers "how long does an A/B test actually need to run to get a result?" You don't need any statistics background — just enter three numbers to get an estimate. This is a Pro-only feature.
The three inputs
- Average daily visitors: Daily visitors to the whole page being tested
- Number of test patterns: The total number of variants being compared, including the control (existing pattern)
- Average conversion rate: Your current actual figure (%)
Reading the results
Assuming the real difference turns out to be 5%, 10%, 20%, or 30%, the calculator shows "time until the result is known" and "total visitors needed" for each. Smaller differences take longer to confirm; larger differences can be surprisingly quick to detect. A small note below the visitor count shows the per-pattern breakdown too.
How the calculation works internally
It uses the industry-standard two-sample proportion test sample size formula (95% confidence, 80% power). It's been cross-checked against Adobe's Target Sample Size Calculator and produces nearly identical numbers.