A/B Test Duration Calculator
Find out exactly how many visitors you need and how long your split test should run to reach statistical significance. Enter your numbers below — results update instantly.
Calculate Your Test Duration
Your Results
Interpretation
Enter your details and click Calculate to see your results.
How to Use This Calculator
Get accurate test duration estimates in four steps.
Enter Your Baseline
Type in your current conversion rate. Find this in your analytics under Goals or E-Commerce.
Set Your MDE
Choose the smallest improvement worth detecting. Smaller effects require more traffic and time.
Add Your Traffic
Enter the average daily unique visitors to the page you plan to test.
Read Your Duration
The calculator tells you how many days to run the test and the total visitors you need.
Why A/B Test Duration Matters
Running a test for too short or too long wastes time and money. Get the duration right and you make confident decisions backed by real data.
- Stop tests too early and you act on noise, not signal
- Run them too long and you delay revenue-generating changes
- Correct sample sizes eliminate false positives
- Proper duration captures weekly traffic patterns
- Data-backed timelines align teams around realistic deadlines
What Affects Your Test Duration
Six variables determine how long your A/B test needs to run.
Baseline Conversion Rate
Higher baseline rates produce smaller variance, meaning you need fewer visitors to detect an effect.
Minimum Detectable Effect
Trying to find a 2% lift takes far more traffic than finding a 20% lift. Match your MDE to realistic expectations.
Traffic Volume
More daily visitors means faster tests. Low-traffic sites should test bold, high-impact changes.
Significance Level
Higher confidence (99% vs 95%) reduces false positive risk but extends the required sample size.
Statistical Power
Higher power catches more true effects. Moving from 80% to 90% power adds roughly 30% more required visitors.
Number of Variations
Each extra variation splits traffic further. Four variations need roughly double the time of a simple two-way test.
Average Test Durations by Traffic Level
How long typical A/B tests take at 95% significance, 80% power, and 10% MDE.
| Daily Visitors | 2% CVR | 5% CVR | 10% CVR |
|---|
Based on a two-tailed test with two variations at 95% significance and 80% power. Your results may vary.
Frequently Asked Questions
Most tests need at least 2 full business cycles — typically 2 to 4 weeks. Never stop a test early just because one variant looks better. You need enough data for statistical significance, and you need to capture weekday and weekend traffic patterns.
MDE is the smallest improvement you want the test to reliably detect. A 5% relative MDE means you are looking for at least a 5% lift over your current conversion rate. Smaller effects need bigger sample sizes and longer tests.
The industry standard is 95%, meaning there is only a 5% chance of a false positive. If the cost of a wrong decision is high, use 99%. For low-risk tests where speed matters more, 90% can work.
Power is the probability of detecting a real effect when one exists. At 80% power, you have a 20% chance of missing a true winner. Higher power (90%) catches more real effects but needs larger sample sizes.
Yes, but you will need to test bigger changes. With 500 daily visitors, detecting a 2% lift could take months. Focus on bold changes — new headlines, layouts, or offers — that drive larger measurable effects.
You risk a false positive. Early results are noisy and unreliable. A variant that looks like a 30% winner on Day 3 often regresses to the mean. Commit to the full duration this calculator recommends.
For standard A/B tests, yes. Testing one variable isolates the cause of any lift. If you want to test multiple variables simultaneously, use multivariate testing — but that requires significantly more traffic.
Each additional variation splits your traffic further. A test with 4 variations takes roughly twice as long as a simple A/B test with 2 variations because each variant needs the same minimum sample size.
Need Help Running Your A/B Tests?
We design, build, and analyze split tests for e-commerce and SaaS teams. Get a free audit and see where your biggest conversion wins are hiding.
- Statistically valid test design
- Full implementation and QA
- Clear reporting with next-step recommendations