3 Replies Latest reply on Jan 5, 2018 10:35 AM by JD Nelson

    Data Normalization set up

    Yun Bai

      Hello Marketo Experts,


      This story began with Marketo warning messages from Notification tab that shows leads won't able to sync to SFDC because of: no country, invalid country, no lead source etc.


      Here I am trying to set up SCs to capture bad data that unable sync lead to SFDC. After captured bad data, I will also need to set up SCs to fix or prevent those issues happen. I am thinking that will either set up batch campaigns and schedule a run time (weekly or so), or create smart lists to generate leads who match the criteria. However, I've been debating which approach would be better to use. If I use Smart Lists, I will have less assets to create, seems simple and easy. While, if I set up batch campaigns, I will need to create SCs, static lists, so more assets will be created which will occupy space in Marketo.


      I did searched for similar topics in the community first. However, I still wasn't sure if both approach will get the same results, or one is better/more accurate that the other one. What would you suggest or to do? Please also suggest or share any data normalization you may know:)


      This is my smart lists looks like:

      No Country:

      No Email address

      No SFDC ID



      Thank you in advance,



        • Re: Data Normalization set up
          JD Nelson

          The main thing to take into consideration is what is called "race conditions" -- if there's any data append or ancillary programs that could update data, you don't want to check until after that. I have a data append service but I still run leads through it after that, so I use a lead trigger based on when that is finished. However, I put ALL my criteria into the same SC so that it can run through a few flow steps at a time instead of 3+ SCs as you've mentioned here.


          I would do a 'lead is created' trigger with "Country is Empty" OR "Email Address is Empty" OR "SFDC Lead ID is empty". Then, on your flow steps you can look at each and populate them all at once (fewer race conditions).

          Screenshot 2018-01-04 09.03.38.png


          You could ALSO run a nightly batch to make sure everything ran through (if after a week or two you notice no leads, then you know your trigger campaign works)

          1 of 1 people found this helpful
            • Re: Data Normalization set up
              Yun Bai

              Thank you JD, especially share with me the "race conditions". This is really helpful.


              The first step for me is to set up and capture leads that missing certain important value. So I am afraid that I have to create smart list or SCs separately. As I do want to have an idea that how many in total are missing value A, or B or C separately, then to dig out the root cause that why there are missing values...


              Your approach will be helpful when try to improve data integrity/quality, such as to pass defined values to fields were empty.  


              Thank you,