I have several long csv files with customers emails loaded in there, if I import them I see several fields unmapped, is it safe to remove it? Id love to keep them but idk how to handle it.
Best answer by retention
TLDR; It’s safe to leave fields unmapped, and it’s actually a good thing.
It’s totally safe to leave Fields Unmapped - basically, it’s as if that column didn’t even exist in your CSV. In most cases, this is what I prefer to do until I have planned to introduce a new “custom profile property” into Klaviyo profiles. When you do, you might want to be more thoughtful of the field name and the possible field values.
For example: I’ve seen many many Klaviyo accounts where they have something like a “Gender” field and the possible values are “Female, Women, Woman, F, W…” where it all implies the gender of their subscribers. Or they have multiple fields like “Gender, Sex, g, ...”
Just remember, it’s always easier to add data to profiles by importing the same contacts again with the new data. If you do, the best practice is to have some intention of what those data will be and keep them as consistent as possible.
Using the example above, this prevents you from having to create overly complicated Segments because the data integrity was not consistent.
For example: Assume I want to create a Segment definition for “Woman” using the above hypothetical data field/value nightmare scenario. It will look like this:
Subscribers WHERE custom profile properties are:
- Gender = Female
- OR Gender = Woman
- OR Gender = F
- OR Gender = W
- OR Sex = Female
- OR Sex = Woman
- OR Sex = F
- OR Sex = W
- OR g = Female
- OR g = Woman
- OR g = F
- OR g = W
If you don’t plan your data fields and data values well ahead of time, that could be what you would have to do every time you want to build a Segment for the “gender!”
Note, it’s not always possible to avoid those scenarios, but the places where I see this data pollution happening the most is when people upload CSVs and just create new field names (without mapping) or just haphazardly mapping data to existing columns.