At the start of the test
Posted: Wed Jan 29, 2025 4:33 am
B) Traffic distribution settings
298-9.jpg
, you decide to play it safe and set the traffic split to 90/10. After a while, you see that the change hasn't made a noticeable difference in conversion, so you move the slider back to 50/50. But returning visitors are still in their original group, so you end up with a situation where the "before change" version has a higher proportion of returning visitors who are likely to convert. Things get complicated very quickly, and the only easy way to get reliable data is to look at new and returning visitors south africa consumer email list separately. However, this will take longer to get meaningful results. And even if both subgroups show meaningful results, what if one of them actually generates more returning visitors? In general, you don't want to do this and change the traffic split during the test.
B) Planning
It seems obvious, but you shouldn't compare data collected at the same time of day with data collected throughout the day or at other times of day. If you want to test for a specific time of day, you have two options.
1. Process your visitors' requests as usual throughout the day, but show them the original version of the page at times of day that you are not interested in.
2. Compare apples to apples – If you are only looking at the first half of the day's change data, compare it to the original first half of the day's change data.
Hopefully some of this will be useful for your conversion optimization . If you have any tips of your own, please share them in the comments.
298-9.jpg
, you decide to play it safe and set the traffic split to 90/10. After a while, you see that the change hasn't made a noticeable difference in conversion, so you move the slider back to 50/50. But returning visitors are still in their original group, so you end up with a situation where the "before change" version has a higher proportion of returning visitors who are likely to convert. Things get complicated very quickly, and the only easy way to get reliable data is to look at new and returning visitors south africa consumer email list separately. However, this will take longer to get meaningful results. And even if both subgroups show meaningful results, what if one of them actually generates more returning visitors? In general, you don't want to do this and change the traffic split during the test.
B) Planning
It seems obvious, but you shouldn't compare data collected at the same time of day with data collected throughout the day or at other times of day. If you want to test for a specific time of day, you have two options.
1. Process your visitors' requests as usual throughout the day, but show them the original version of the page at times of day that you are not interested in.
2. Compare apples to apples – If you are only looking at the first half of the day's change data, compare it to the original first half of the day's change data.
Hopefully some of this will be useful for your conversion optimization . If you have any tips of your own, please share them in the comments.