Comment utiliser : A/B Test Significance Calculator
- Enter the visitors and conversions for version A, your control.
- Enter the same for version B.
- Pick a confidence level. 95% is the usual choice.
- Read the result, then check the p-value and the range of the difference.
What statistical significance means
Random variation means two identical pages rarely show identical rates. A significance test asks how likely a difference this large would be if the versions performed the same. That chance is the p-value. At 95% confidence, a result counts as significant when the p-value is below 0.05.
How it’s calculated
This tool uses a two-sided, two-proportion z-test. The formula is z = (rate B − rate A) ÷ √(p(1 − p)(1/visitors A + 1/visitors B)), where p is the combined conversion rate.
Avoid common mistakes
- Don’t stop early. Checking daily and stopping at the first significant result produces false winners.
- Plan your sample. Use the A/B Test Sample Size Calculator before you start.
- Run full weeks. Weekday and weekend visitors behave differently.
Aussi recherché comme : ab test calculator, ab test significance calculator, statistical significance calculator, split test calculator, ab test p value, conversion significance.
Outils complémentaires : A/B Test Significance Calculator
- CTR Calculator — Enter clicks and impressions to get your click-through rate, for ads, email or Search Console.
- ROAS Calculator — See what each ad dollar brings back.
- UTM Builder — Add utm_source, utm_medium, utm_campaign and friends to any link, see which GA4 channel the visit will land in, and tag a whole list of URLs at once.
- Calculateur de pourcentage — Sept calculs de pourcentage courants sur une seule page.
Questions fréquentes
Is anything I enter saved or sent?
No. The calculation runs in your browser. Your last inputs are remembered on this device only, so the page opens where you left off.
What p-value is good for an A/B test?
Below 0.05 is the usual standard, which matches 95% confidence. Use 0.01 for high-stakes changes.
Is 95% confidence the chance that B is better?
Not exactly. It means that if there were no real difference, a result this extreme would appear less than 5% of the time.
Why is my result not significant?
Usually the sample is too small for the size of the difference. Keep the test running to your planned sample size.
Puis-je installer A/B Test Significance Calculator en tant qu’application web et l’utiliser hors ligne ?
Oui. A/B Test Significance Calculator est une application web installable (PWA) : elle s’installe depuis votre navigateur, sans passer par une boutique d’applications. Cliquez sur « Installer l’application » en haut de cette page dans Chrome ou Edge (Windows, macOS, Linux, ChromeOS) ; sur Android, utilisez « Partager » → « Ajouter à l’écran d’accueil » sur iPhone et iPad ; ou « Fichier » → « Ajouter au Dock » dans Safari 17+ sur Mac. L’application s’ouvre alors dans sa propre fenêtre avec sa propre icône, et comme elle s’exécute entièrement dans votre navigateur, elle continue de fonctionner hors ligne. Firefox sur ordinateur de bureau ne permet pas d’installer d’applications web. Comment fonctionne l’installation.