Comment by SeanDav

Not sure how you can get much of value from an A/B test with multiple changes, especially if one is claiming that only 1 of those changes is what is responsible for all the improvement.