Hi, we've been running test in Test&Target containing 5 experiences. Once the experience is shown to visitor, we also send evar to analytics, that helps to cross check Test & Target data. From the beginning of the test T&T balanced traffic between experiences equally, but after few days, it started to prefer one /first/ experience. We might made any mistake in setup /maybe we should change the test type from "landing page campaign" to "A/B..N campaign" - see screenshots/. Any help or idea would be appreciated. Thanks
Yes, change it to an A/B campaign from a Landing Page campaign. Also remove your experience level target or recipe level targeting and simply apply that condition at the location or mbox level.
Because visitors are qualifying for that first experience (because the targeting is there), is why you are getting so much traffic for Experience A.
Either reset the data and start again but better to make a copy, change to A/B and relaunch so you have this data if you ever want it later.
that makes sense. Next time we'll apply target on mbox level. Thank you!
I'd have to look at the test setup but I imagine it is visitor requalifying for the test and since they qualify for the first experience they are sent there. Think of it as a top to bottom approach. As they re-qualify, Target will start with the targeting from Experience A and see if visitors meet that condition. If not it moves on to Experience B, etc....
Copy the test, change to A/B, and remove Experience Level target and apply it to Location level and you are home.
Brian thank you very much for your answer. Could you please be more specific on this: "Because visitors are qualifying for that first experience (because the targeting is there), is why you are getting so much traffic for Experience A. " From my point of view, all experiences are equal because all experiences have exactly the same targeting /excluding visitors with Opera browser/. Therefore I still didn't get why all experiences are not equally served to visitors but first one is more preferred.