Workflow help - Duplication

[Alteryxuserhere]

Hi there

I have the following data that has been cleaned and has some duplication that I would like to keep (e.g. row 1 and row 6 and 7), because although the email and mobile number is the same they have different information in the other columns in Product and Channel. 

I would like to count the number of times they appear per channel and have used the summarize tool as shown below and would like to combine this new information back into the orginal data set as new collumns and have tried to use the Join to to join back via email address however its creating alot of additional duplication. Can someone please suggest how I can do this without all this duplication?

 

[gawa]

hi @Alteryxuserhere 

You can simply Summarize by 3 fields; Channel, Email Address, Mobile Number. And, join to the original data by all these 3 fields.

Please have a look at attached WF.

[Alteryxuserhere]

@gawa thanks for your reply

but if I have to count the email and mobile number separately I cant group them into summarize as if the same customer has the same email address but a different mobile number or different mobile and same email address, it won't count properly. This is why I have separated them into two separate summarize groups in the workflow however I'm unsure how to remove the duplicates.

[gawa]

@Alteryxuserhere 

Should it be like this?

[Alteryxuserhere]

@gawa thanks for the reply. Can you explain to me how the duplication occurred in my workflow and how the combination of the channel + email/mobile reduce the duplication?

[gawa · Accepted solution]

@Alteryxuserhere My observation is as follows. Please note that I'm not fully understanding the context of your dataset, this advice is just for your reference. If you wan to learn more on JOIN concept, please go to interactive lesson in ACADEMY.

1) At JOIN(7), thought you specified join key only "Channel" but it should be "Channel" and "Mobile Number". That generated unnecessary 3 records as marked in red below picture.(Mobile number is different but joined)

2) At JOIN(3), thought you specified join key only "Email Address" but it should be "Channel" and "Email Address". That generated unnecessary 3 records as marked in red below picture.(Channel is different but joined)

3) Though you JOIN two data from inner join of JOIN(3) and JOIN(7), but it should be joined in series like I did in my WF shared in the above post.