(Real life struggle) A/B Testing for measuring the effects of different touchpoints on customers

So we are colaborating with the data analytics department at our company and the business came up with the following hypothesis: "Customers who are engaged through the website have a better experience and hence a better turn-over against the ones that don't engage through the website" with the intention to prove that "pure digital" customers are more efficient at creating revenue against the "classically engaged" customers.

I am helping the analytics guys gather data for this but I just don't …

a/b testing analytics b testing business customers data data analytics datascience effects experience hypothesis life testing through website