Expand my Community achievements bar.

Join us for the Adobe Journey Optimizer Community Q&A Coffee Break on 28th February at 8 am PT with Ariel Sultan, Daniel Wright, Josee Haeley and Rohit Basuri.
SOLVED

How are identical fields matched between two different datasets in CJA

Avatar

Community Advisor

While configuring the connections in CJA, I am adding multiple datasets. And for some strange reason, I see some field overlaps get identified correctly, but this only happens for some fields. While creating the schema in AEP, do I need to create the fields in a specific manner so that CJA identifies them correctly? I tried keeping the same name, the same type -> but this is not working as expected.

 

Because of this, I cannot combine two datasets and generate a consolidated report. E.g., The brand field is present in two datasets.

 

Am I missing anything, or is this a product bug?

Topics

Topics help categorize Community content and increase your ability to discover relevant content.

1 Accepted Solution

Avatar

Correct answer by
Community Advisor

@Anil_Umachigi, I have connected the two datasets with the unique id (say customer id). I have two brand name fields in the dimensions, one from each dataset (ideally column gets tied together there will be only one combined dimension field).

 

While configuring the connection, if the CJA understands that these columns are the same type, I can easily derive stats combining two datasets.

View solution in original post

6 Replies

Avatar

Community Advisor

@ChetanyaJain Trying to understand the end report you are trying to get to, will you use brand field to unify data from both datasets? 

Also irrespective of what field name or type (every matching names and types) fields from both dataset will be ingested into CJA, these can then be filtered within data view to either include both or either within workspace UI. 

 

Now coming to consolidated report, this will be purely based on the connection ID you have chosen while creating the connection and not individual fields. 

Each of the brands field will be independent dimensions within the UI. 

 

 

Avatar

Community Advisor

@Anil_Umachigi, here's the setup:

  • Created a connection
  • Added three datasets (two events & one lookup)
  • The two events (assume appointments & invoice) have one common field in common - brand id/brand name
  • Look up dataset is brand details - in AEP, it relates to two event datasets.

Here are some observations:

  • In the connection preview, we see OOB timestamp fields overlaps between both the datasets
  • Event Type (oob) shows as two fields
  • Same with Brand Id/Brand name

Since it's shown as two fields, the dimensions we create will be linked with respective datasets.

 

When we drag brand dimension on reports (say table format) - and add metrics from appointments & invoice fields/events -> it doesn't stitch together or create a consolidated report for a brand.

Avatar

Community Advisor

right ok this helps. 

ok how about you try this, use the unique id ( used for connection to stich together the event data sets) in dimension with metrics you need. 

Now break down the unique id by brand and hopefully this should do the trick. 

 

let me know if that helped

 

Avatar

Correct answer by
Community Advisor

@Anil_Umachigi, I have connected the two datasets with the unique id (say customer id). I have two brand name fields in the dimensions, one from each dataset (ideally column gets tied together there will be only one combined dimension field).

 

While configuring the connection, if the CJA understands that these columns are the same type, I can easily derive stats combining two datasets.

Avatar

Community Advisor

@ChetanyaJain Are you sure the same column name will get tied together? i doubt this is the case unless it's a new feature. 

Does it create a new column with stiched brand details?  

 

Avatar

Community Advisor

Currently, it doesn't and I wish we can tie them together. It does tie a few columns - so not sure how it decides which columns to tie... but I definitely wish to tie an event-type column.