You do A/B testing manually or offline from time to time.
You have some data, so you put it into a spreadsheet. Then, you look at the rates of conversions and decide which version worked best.
This is the trap.
The statistical significance method is too hard to understand.
But you can’t be sure of your findings if you don’t have it.
There may not be enough confidence to pick the right winner.
Here’s a free solution: you can use 2 free online tools to work with the data you’ve hand-collected:
1. A Kissmetrics tool
2. Another one by Survey System
Source: getdatadriven.com