Experiment Design for AB Testing

Hello everyone,

We are planning to conduct an A/B test to determine whether the price increase for insensitive products in our product catalogue would lead to a statistically significant difference in the revenues between test and control groups and determine price increase post that.

However, before we conduct the pilot for A/B testing we want to determine what should be the delta that we should expect between sample means (post pilot) at which we can consider the sample means to be statistically significantly different. Has anyone worked on a similar problem? Please let me know and I can share a sample data with further information.

Thanks in advance for your help.



Less is more. Stay pure. Stay poor.
Revenue is a composite outcome and wouldn't describe which items are being purchased, which seems important. It is up to you all to define a situationally import change in revenue, we would have not idea what would matter.