[Target] Customer attributes - 1 to many relationship management

[chamito]

Hi Community,

we are in the process of integrating Customer Attributes into the People function of the Adobe Experience Platform.

As all you can imagine, we want to fill in the mapping fields with useful data related to the users; it happens that in the CRM we're working on, one user record is related to many records on another table, for example each unique user record on table USERS is related to multiple user interest categories on table INTERESTS.

Our question is: would Adobe Customer Attributes manage multiple lines like:

User ID	Interest
user1	interest1
user1	interest2

or, as I guess, the only way to manage this will be serializing the values like this?

User ID	Interests
user1	"interest1;interest2"
user2	"interest1;interest3;interest4"

Thanks in advaance,

Daniele

[Gokul_Agiwal · Accepted solution]

Hi @chamito 

Yes, you're right.  the second approach is correct.   I've used the 2nd approach only in my project. 

See the below example which I'm using in my project. 

User ID	Data
user1	"country_name, state_name, product_name, preference, interest1 | interest2 | interest 5, dept_name, job_role"
user2	"country_name, state_name, product_name, preference, interest1 | interest3 | interest4, dept_name, job_role"

Actually Customer attributes works as profile level attributes which is always in key value pair where each key as user ID has only one record. It's not possible to handle one to many relationship here with Customer attribute. 

Another point 

• In case, If you try to upload multiple rows with the same User ID, the last row will overwrite the previous ones — only the last one remains.

• Therefore, serialization is the only way here. 

Hope this helps. 

Thanks

[chamito]

Hi Gokul,

thank you very much for the quick answer, that's very clear.

On this basis, I try to ask a follow up question: to better manage resource quotas on the CRM platform, is it possible to split the export in different chunks?

According to documentation (https://experienceleague.adobe.com/en/docs/core-services/interface/services/customer-attributes/crs-... it should be possible within this two conditions:

• each chunk elaboration has to be completed before uploading the next chunk
• records in each chunk have to include different user ids to not overlap with previous

To sum up, I'm imagining a scenario like this:

• DAY 1:
  • 06:00am: insert record 1-10000
  • 12:00pm: insert record 10001-20000
  • 06:00pm: insert record 20001-30000
  • 12:00am: insert record 30001-40000
• DAY 2:
  • 06:00am: update record 1-10000
  • 12:00pm: update record 10001-20000
  • 06:00pm: update record 20001-30000
  • 12:00am: update record 30001-40000
• and so on

Is that a realistic scendario?

Kind regards,

Daniele

[Gokul_Agiwal]

Hi @chamito  Daniele, 

Spot on, Yes the scenario mentioned above is correct and aligned.  Each chunk fully processed by Adobe server before you upload the next one. 

Please read all the guardrails here,  Data File and Data Sources | Adobe Experience Cloud 

In my case, I'm using FTP setup which upload the data off business hours and processed it. 

Feel free to post/ ask if any more questions. 

Thanks