Sample sizes for testing on mid-traffic websites
Sample sizes for testing on mid-traffic websites When setting up an AB test (or split of MVT), getting enough traffic to significantly conclude a result can be very hard. A lot of the online calculators tell me I need to get 1000s of sessions on each variant before I can even think about drawing any conclusions – e.g. AB Tasty tells me I need 7k sessions for each variant (based on 5% CVR and a…