When importing contacts with a column of contacts they have a relationship to, both the contacts being imported and the contacts they have a relationship with, are added to the group selected for the imported contacts to be added to.

We created a new organisation sub-type, called Electorate. We created a relationship between an Individual and Electorate of "Candidate for"

We imported a bunch of electorates as Electorate type contacts.

We imported a group of individuals with a column that has the Electorate that they are a Candidate for, and used the "Candidate for" matching to "Organisation Name (match to contact) *"

When we did the import we said to add to a group called Candidates. The import did not recognise the existing Electorates and rather created a whole new set of contacts (with the correct Electorate type)" and then added both the Individuals and the Electorates to the Candidates Group (obviously only the individuals we were importing were intended to be added to that group, not the contacts with which they had a relationship. When you remove the mistakenly added contacts from the Group, there is a history that they were in the wrong group, that can't be deleted through the UI.

Only the contacts being imported in the csv should be added to the selected Group, not the contacts named in the csv for which a relationship is being created, or else an option should be provided to "Add secondary contacts to Group" and that could be another list of groups so that the contacts that have a relationship with the imported contact can be added to a different group.