A/B Test Calculator
How many visitors/conversions do you need for a statistically significant A/B test result? Calculates sample size per variant and estimated duration based on current rate + expected uplift + confidence level.
Both groups combined: 90.754. Each variant needs to generate ~1.589 conversions for the test to have a statistical basis.
With 1.000 visitors/day, this is the time needed to reach sample size. Consider running at least 1 full weekly cycle (weekends behave differently).
Before testing: what has your competitor already validated?
Running an A/B test costs visitors and time. Batedor shows you what type of campaign your competitors have been running consistently, a signal that something works. Start testing what the market has already proven, not in the dark. 14-day free trial.
How it works
How to calculate A/B test sample size
The calculator takes your store's current conversion rate, the uplift you want to detect (the minimum detectable effect, or MDE) and the confidence level (usually 95%) to estimate how many visitors each variant needs before the result becomes statistically significant. The smaller the uplift you want to catch, the more traffic the test demands. From your daily traffic it also projects the estimated duration in days.
Example: your page converts at 2% and you want to detect a 10% relative lift (from 2% to 2.2%) at 95% confidence. The tool points to roughly 30,000 to 40,000 visitors per variant. If the store gets 2,000 visits/day split between the two versions, the test runs around 30 days to reach the full sample.
A common pitfall in Brazilian e-commerce is stopping the test early, on the first day B looks like it is winning. Wait for the full sample and run at least one or two whole weeks, covering weekdays and weekends, because buying behavior shifts a lot between Monday and Saturday and around dates like Black Friday.
Learn more
Articles to put this tool to work
KPIs de Inteligência Competitiva: O Que Medir para Não Monitorar no Vazio (2026)
Monitorar sem indicador é coletar ruído. Os KPIs que transformam observação de concorrentes em sinal acionável.
Como Transformar Dados de Concorrentes em Decisão Comercial
Coletar dado é fácil. Tomar decisão é difícil. Aprenda a transformar monitoramento em ação.
FAQ
Frequently asked questions
- What is an A/B test calculator?
- It's a tool that calculates how many visitors and conversions you need per variant for an A/B test to reach statistical significance. From your current conversion rate, expected uplift and confidence level, it estimates the required sample size and test duration. Runs free in your browser.
- How do I calculate the sample size for an A/B test?
- Enter your current conversion rate (baseline), the minimum uplift you want to detect and a confidence level (e.g. 95%). The calculator returns the number of visitors and conversions needed per variant for a statistically significant result, so you don't have to run the statistical-power math by hand.
- How long should an A/B test run?
- Duration depends on the required sample size divided by your traffic volume. The tool estimates how long the test should run based on the calculated sample and your usual number of visitors, so you don't stop the test too early and draw a premature conclusion.
- What is the difference between confidence level and expected uplift?
- Confidence level (e.g. 95%) measures how likely the result isn't due to chance. Expected uplift is the minimum improvement you want to detect between variants (e.g. going from 2% to 2.4% conversion). The higher the confidence and the smaller the uplift, the larger the sample you need.
- Is the A/B test calculator free?
- Yes. Batedor's A/B test calculator is 100% free, runs right in your browser and needs no signup. Adjust the current rate, uplift and confidence level and instantly see the sample size per variant and the estimated test duration.
