Harmonisation in V2.0 Data Studio

Carolyn

Hi

Trying to understand the best way to get a better harmonised record, when you have multiple lines for a customer.

I am finding if you take merge best value and use most common l lose some data

For eg - if you have:

Home Phone work Phone mobile

9999 2222 9999 2222

It only picks up one of the phones when it supplied the harmonised line

If you had

Home Phone work Phone mobile

9999 2222 8888 1111

This would pick up each phone differently

Then l took total length instead of most common value and this takes the longest length of each column and if all the same length, seems to take the first record

What l am wanting to do is take the most common value for each field and return this in the field.

Eg

Name surname address suburb pcode DOB Home ph Work ph

John Smith 22 Smith Street Ballarat 3220 01/01/1970 8888 1111 9999 3333

Johnn smith 22 Smith street Ballarat 3220 01/01/1970 9999 3333 8888 1111

John Smith 22 smith street Ballarat 3110 02/01/1970 9999 3333 8888 1111

i would like harmonised record returned to be the most common result for each field value - so based on above:

John Smith 22 Smith street Ballarat 3220 01/01/1970 9999 3333 8888 1111

I played with the additional options with no luck.

Have you got any suggestions.

thanks

Carolyn

Henry Simms

Hi @Carolyn , I used your data with the same Harmonize step settings:

• Record Processing: Merge best values
• Value Priority: Most common

I seem to get the result you wanted (3 input records in the top table, result of the harmonize step in the bottom table):

What result are you actually seeing with your example data?