Solved

Index for Wave based Communications

  • 26 April 2018
  • 6 replies
  • 52 views

We've been working this on a while and can't seem to come up with an easy solution.



I have a list of over 50k emails I need to contact from an unindexed MDA object. Besides their name, these emails are linked to a specific account name, and an SFDC ID, but nothing else much to make them unique in a way that's meaningful.



I want to comm all of them, but not at the same time. There doesn't seem to be an easy way to segment them into say, groups of 5000, and Support was not able to point me in the right direction.



We can probably go back to the base level data source in our company and use its index, but that means there's no randomization, and I segment my customers by tenure, which is not ideal. Plus I'd need to work in a couple weeks of time for R&D to expose that data to us.



I'd love to hear solutions anyone has come up with, especially if it still allows us to use Advanced Outreaches as our solution.
icon

Best answer by dan_ahrens 12 October 2018, 17:37

View original

6 replies

Badge +3
Random idea (see what I did there?): create a new column and use the string functions to populate it with the third character of the account name (or the first character, or the secibd character of the address, or whatever). You can then create segments by selecting the records with that column = "a" or "b" or "c", etc. 
Userlevel 7
Badge +1
Hi Brad, I recently started going through all of the pending questions . Was checking to see if this has been answered. 



Please let us know if you need any help here?
Absolutely. I still don't believe this has been solved because we can't apply a straight numbered indexed to our data sets.
Userlevel 7
Badge +1
Hi Brad, Did you try with @karl suggestions ?
No, as that solution is not what we're looking for and doesn't result in even segments. We still have to export our data out of gainsight, manually wave via excel or rstudio and then import back in.
Userlevel 7
Badge +2
Thinking through this use case, Brad, I believe the best solution would be to export the list to Excel and then add a column for your data using the =RANDBETWEEN function in Excel to assign a randomized cohort group to each record. While not precisely even segments, it will be close to that, but fully random. Then pull that data back into MDA.



Excel has a pretty comprehensive library of formula fields - 30+ years of development and investment will have that effect 🙂. Sometimes it's just easier to manipulate the data in Excel.

Reply