
Did B actually win?
The new button got 6.0% and the old one got 5.0%. Is that real, or is it Tuesday? This calculator answers it with the same two-proportion z-test the big testing tools use, in a spreadsheet you can read.
Works in Google Sheets and Excel.
FIVE INPUTS
Visitors and conversions for A, visitors and conversions for B, and your confidence level.
WHAT COMES OUT
- Verdict: B WINS, A WINS, or NOT SIGNIFICANT YET
- P-value (two-tailed), both conversion rates, relative lift, z-score
- The visitors per variant you needed to confirm this lift at 80% power, so you know whether to trust it or keep running
THE PLANNING TABLE
How many visitors per variant you need to detect a 5, 10, 15, 20, 30 or 50 percent lift from your current rate. Decide the sample size before the test, not after.
What you get
- The calculator (.xlsx, opens in Google Sheets or Excel)
- A 1-page quick-start guide
Who it's for
Marketers, product managers, founders and indie hackers running tests without a data scientist to ask.
Good to know
- Formulas are visible, not hidden. Check the maths if you like.
- No macros, add-ons or sign-ups. Works in Google Sheets (free) and Excel 2016 or newer.
- Instant download.
Need a custom app built for your business? I build native iOS apps and web tools from prototype to launch. Services and pricing · LinkedIn
Indie dev and launch
A/B Test Significance Calculator
Google Sheets + Excel
Visitors and conversions for A and B in, verdict out: p-value, lift, and how many visitors you actually need. Plus a sample size table.
One-time price in US dollars, checkout by Stripe. After paying you land on a page with your download link. Each link works for 15 minutes, and reopening that page makes fresh ones.
- AB-Test-Significance-Calculator.zip