CiviCRM Community Forums (archive)

I have been working through an import process and finding some counter intuitive things.  Not sure that this is a bug report, just unexpected stuff.

When importing, civicrm always seems to use the "strict" dedupe rule.  In postings here, you guys said that when the admin is working, the "fuzzy" rule is always used.  I haven't seen a setting to control this, maybe I missed it.  If the strict rule is less likely to find a match, then on import civicrm is more likely to create a dupe.  When I broadened the strict dedupe rule (fewer criteria to meet) then civicrm creates many fewer duplicate contacts. 

To make this work, I now have a strict rule that is fuzzier than my fuzzy rule!

The strict dedupe rule was set to match only on email.  Not sure if that was the default or if someone here changed it previously.  Still that made no sense when importing contribution contacts where we only have check data -- first and last name and street address fields.  Using email would make sense for online registrations, but not for this particular import.  Using this rule, 100% of my data created duplicates. 

I can live with it, but maybe a list box on the import page where you can choose the dedupe rule to apply for your data file would be an improvement.  Documentation explaining how it works would be good too.

Thanks.