I'm looking to dedupe some contacts in my org, and we have birthdate data for a lot of our contacts. I'd like to create a scenario where some other things match as well, but importantly that birthdates are not both set to something different. So it's fine if they're both blank, and it's fine if one is blank and the other is set, and it's fine if they both match, but it's not fine if they're both set and they don't match.
Your first scenario will have a condition on Fields to Compare screen that includes birthday. You can use the date comparison type, but make sure to check the "Blank Values" algorithm. This allows for two blank fields to be considered a match.
This will cover those birthday fields that have the same value and those that have no value at all.
Your second scenario will not have the condition of birthday in the Fields to Compare. Instead, you're going to go to "Filter Match Results" and condition on the birthday and say it equals blank. By doing this, you're saying that at least one record in every match will have a blank value.
You have to be careful with this one, as you may get groups of more than two where two out of the three records may have a date for birthday (and they may not match). The safest way to do these automatically is to do a very simple winner rules tha says the record with a date in the birthday field wins. Make sure to use a score delta of 1 or higher.
If there is more than one record with a date, it would be a tie and no winner will be chosen and it will not process until you choose a winner manually.
If you think this is a very small batch, you could no apply a winner and just review each one manually yourself and choose the winners. You would then also deselect any losing records you don't want to be included in the merge.
Your first scenario will have a condition on Fields to Compare screen that includes birthday. You can use the date comparison type, but make sure to check the "Blank Values" algorithm. This allows for two blank fields to be considered a match.
This will cover those birthday fields that have the same value and those that have no value at all.
Your second scenario will not have the condition of birthday in the Fields to Compare. Instead, you're going to go to "Filter Match Results" and condition on the birthday and say it equals blank. By doing this, you're saying that at least one record in every match will have a blank value.
You have to be careful with this one, as you may get groups of more than two where two out of the three records may have a date for birthday (and they may not match). The safest way to do these automatically is to do a very simple winner rules tha says the record with a date in the birthday field wins. Make sure to use a score delta of 1 or higher.
If there is more than one record with a date, it would be a tie and no winner will be chosen and it will not process until you choose a winner manually.
If you think this is a very small batch, you could no apply a winner and just review each one manually yourself and choose the winners. You would then also deselect any losing records you don't want to be included in the merge.
