Use the Calculator below to figure out the necessary sample size and test duration to ensure your A/B test results aren’t random.
Your current conversion rate can be figured by dividing your last month’s conversions by the total visitors of the last month.
Example: If you made $400 and had 2000 visitors your equation would look like this.
400 ÷ 2000 = .2 or 20%
It’s not important this number be over-thought; take an educated guess.
If you’re a pro and working for a client – you may expect to improve 20%. If you’re starting out maybe 5-15% is more appropriate. Use your best judgement.
What this calculator tells you is two things, for two different confidence levels.
- Visitors required for accurate test
- Number of Days to run test
Confidence Level is the likelihood that your test results signify true improvement or loss, rather than being a random trend.
80% means 80% chance that the variants performance is ‘true’.
95% means 95% chance that your test results are ‘true’.