Upload audiences into Adobe Experience Platform

[LSAndrade]

Recently I tried to upload a custom audience. 

I uploaded a simple CSV file containing 3 columns. The first column was a identitie and the others are just additional data that I need on my profiles.

The file had 30k lines and each of them is a unique identity. The dataset created from this process has the same amount of records ingested. So I assume everything went all right with the upload process. But the audience had only 3.35K profiles. 

I open a ticket and the support mention that the dataset should be marked as profile. If this information is right I need to upload a external audience I should upload two times the same file just to had all profiles in the audience? This tutorial don't mention anything about set the dataset to profile too: https://experienceleague.adobe.com/en/docs/platform-learn/tutorials/audiences/upload-audiences

[RiteshY18]

i think Adobe support is right, you should check DataSet/Schema where file has been uploaded to and given link gives a high level of information not end to end ingestion process 

[LSAndrade]

Just to clarify I use the option "Import audience CSV" on Workflows menu. As in this tutorial: 

https://experienceleague.adobe.com/en/docs/platform-learn/tutorials/audiences/upload-audiences

When use "Import audience CSV" I'm uploading a externally managed audience. I'm not uploading data to dataset, this is not profile fragments. All the datasets created are created by this process, so are system generated. Still don't make sense that I should manually check for profile a system generated dataset. 

[RiteshY18]

Is dataset enabled for profile, then uploaded audience assed against the identities thus leading profile merge and reducing number of profile. 

On the profile, search for one of you uploaded identify and check identity graph

[Jennifer_Andrews · Accepted solution]

If you are uploading an audience then, no, dataset/schema do not need to be profile enabled.   

Custom upload audiences work differently and have a backend process that ingests data into data lake. You will see on your Audience a dataset/schema and profile batch. This profile batch is ingested into a system-defined dataset.