Designing the new deduplication feature

by Andrew21. August 2013 1:29

Contact databases often suffer from a common issue - duplicate contacts.  They creep in, they are imported in and are made all the time through inconsistent naming or incomplete information.  No one is really safe from the pain of duplicate contacts and it can cost a company significant embarrassment when duplicate contacts are both treated as the actual sole record. 

To date we have not had a good solution for dealing with duplicates beyond manual intervention. So I am pleased to announce the next major feature to be added to saleslifecycle - de-duplication!

Detecting Duplicates

It all starts with detecting if a duplicate contact record exists. Our method is a rather brute force approach of comparing all of your active contacts with each other.  Matching contacts are made by comparing their company name, last name, first name and primary email address and measuring the similarities.

From this we build a probability score out of 100 to determine if they are in fact duplicate records.

For example:

John Smith at ComWire International
Jonathan Smith at Commerc International

Would be given a probability of around 70% of being a duplicate.

If the score is greater than 80% we will add a warning on the contact page of a possible duplicate contact.  Contacts with a possible duplicate will be searchable via the new advanced search option 'Has Duplicate'

Resolving Duplicates

If you believe that the contact records are not duplicates then you can choose to 'ignore' the match and it will not be suggested again.  If you believe that they are duplicates then you have two choices: Merge or Delete one of them.

When will this be available?

The final design is being worked on now and we expect to see this new feature rolling out in the next week or so.

Until next time.

Tags: ,


Comments are closed
1st user free
1000 credits

Month List

Feed Subscribe