Resolve imbalance visitors

[anushap94253501]

How to resolve imbalance visitors to all experience in A/B Testing. Thanks and Advance, Anusha prabaharan.

[ParitMittal]

Hi Anusha,

Thanks for reaching out to Adobe Target Community.

We are unable to understand your query completely. Hence can you please elaborate your query in detail so that we can further help you with the same.

Regards

Parit Mittal

[anushap94253501]

Hi ParitMittal,

I have a test splited with control / Recipe 1 / Recipe 2.(30% - 30% - 30%) Visitors report combined with Adobe Analytics.  Comparing to control, Recipe 1 and 2 showing  more number of visitors. Imbalace of visitors. I have attached the image. Request you to help to find rootcause of the problem. Hope this info helps. Thanks in Advance.

[ParitMittal · Accepted solution]

Hi Anusha,

Are you sure it is (30%,30%,30%) ?. As the The total of the percentages for the test should add up to 100%.

Also,If your activity has multiple experiences that are set up with specific percentage targets, for example four experiences each configured to receive 25% of the traffic, a random number is generated and that number determines the experience shown to a visitor. In this example, if the random number is 30, the visitor would receive the second experience because 30 is between 25 and 50, the second range when dividing 100 into 25% chunks.If there has been very little traffic, for example at the beginning of a campaign, the traffic might not be split exactly as specified in the percentages. The more traffic you get, the more exact the percentages will become.

Regards

Parit Mittal

[laurenr36581018]

We frequently have this problem as well. It's always Experience A that has fewer visitors than the new experiences created, despite the traffic split being equal across the experiences. Right now I have a test with the following results in Analytics:

	Visitors
Exp A	5856
Exp B	6930
Exp C	6913
Exp D	7056

We have submitted various questions to Client Care about this and have not reached any resolution. Our only hope is that this will stop occurring when we migrate over to at.js and use A4T rather than our current mbox.js/Sc-Tnt integration.