VWO Logo VWO Logo
Pricing
Request Demo

A/B Test Sample Size & Duration Calculator

Wondering how long your A/B test should run on VWO? Estimate required duration and sample sizes for different statistical configurations with VWO's free calculator.

A value for this field is required.
Metrics
%
A value for this field is required.
$
A value for this field is required.
%
A value for this field is required.
%
A value for this field is required.
Testing Objective
%
A value for this field is required.

Estimated campaign duration ~ days

Due to seasonality effects, it's recommended to run your test for at least 7 days.
Want to shorten the duration?

Try increasing the target to conclude in 4 weeks or less.

Total visitors required

Required MDE:

To detect a change in conversion rate from 5 % to a target of 5 % , your campaign needs to run for approximately 21 days with your current traffic and statistical settings.

To detect a change in average revenue from 5 % to a target of 5 % , your campaign needs to run for approximately 21 days with your current traffic and statistical settings.

Review and correct errors to calculate total visitors required and campaign duration accurately.

Minimum Detectable Effect (MDE) over time

Adjust your campaign duration using the table below to see how MDE and required visitors change over time.

Note: The calculator is designed for VWO's enhanced SmartStats engine.
To get estimates using the classic stats engine, please use this calculator instead.

Say hello to enhanced SmartStats - our Bayesian-powered sequential testing engine

Get actionable results faster

Get clear insights on whether your variations outperform, underperform, or match the baseline. Our engine automatically recommends disabling underperforming variations to speed up results with fewer visitors. Plus, SmartStats tracks your experiment’s health—monitoring minimum run-time, data tracking, and more—while alerting you to errors for quick corrective action. Ensure reliable, faster results with confidence.

Get clear recommendations from reports

Take complete control of statistical parameters

No more one-size-fits-all experiments. With enhanced SmartStats, you can fine-tune statistical parameters—like Region of Practical Equivalence, Statistical Power, False Positive Rate, and Minimum Detectable Effect—based on your priorities. Set higher precision for critical experiments, such as revenue tests, and lower precision for low-priority ones like design tweaks. Tailor your approach for smarter, more efficient experimentation.

Configure statistical parameters according to your needs

Pick between improvement and non-inferiority modes

Choose your testing objective based on your needs: "strict improvement" for guaranteed uplifts, or "improvement or equivalence" to maintain your guardrails. For accurate, error-free results, use the Fixed horizon approach, which only calculates statistics after the full sample size is reached. Alternatively, the Sequential testing approach adjusts for peeking errors, ensuring your reports are always accurate, no matter when you check them.

Configure metric types that suit your testing objectives

Increase your conversions, starting today.

Start Free Trial Request Demo

Awesome! Your meeting is confirmed for at

Thank you, for sharing your details.