Offline analysis A/B test - sum of revenue squared

[larav34689834]

I downloaded the Excel file from adobe that contains detailed calculations for the analysis of offers when conducting an A/B test (or other tests).

There are three parts: conversion rate calculations, revenue per visitor, and average order value calculations.

For the revenue per visitor calculation, the "Sum Rev Squared" is required.

I can't find a detailed explanation of how this number is calculated.

Am I correct if I take the revenue per visitor, then square this, and take the sum of all visitors?

I assume this value can only be calculated if you have the revenue for each visitor being a part of the offer.

I don't have a way to verify is my calculations are correct or not. It would be great if someone can confirm this or give a correct calculation for this.

Kind regards

[Amelia_Waliany · Accepted solution]

Hi there and thanks for your question! Below is a basic breakdown:

*Experience A*

```Person A: buys 3 products for a total of $15

Person B: buys 1 product for a total of $100

Person C: buys 9 products for a total of $235```

*Experience B*

```Person D: buys 1 product for a total of $150

Person E: buys 3 products for a total of $78

Person F: buys 2 products for a total of $12```

The math is:

Sum of Rev Squared for A:

```(15^2) + (100^2) + (235^2) = 65,450```

Sum of Rev Squared for B:

```(150^2) + (78^2) + (12^2) = 28,728```

Hope this helps!

[larav34689834]

Hi Amelia,

Thanks a lot for this clear answer, it really helps me !

Kind regards,

Lara

[cainej77377828]

Hi -

Is there a way to quickly create the sum of revenue squared as a calculated measure?  Measuring confidence on calculated measures, so getting access to the data to preform this function in Excel - is proving difficult.

Thanks

[melissag7357285]

How do you do this in Workspace?