While CATS will match an email address when someone applies more than once, someone who applies through Indeed and through the career portal will not be matched since Indeed uses an indeedmail address. We would like either a queue of matched names or phone numbers for review; or a pop-up warning of a potential duplicate based on these other criteria when reviewing a candidate.
Yes! Please please add this, we have the same issue. Being able to run a “possible duplicate” report based on names or phone numbers would be incredibly helpful and allow us to keep our data base clean.
This would do bundles for us. The same name and lack of email matches causes problems. There should be a need for some unique id or a possibility of duplication lookup and a fix for the same.
Thanks for the suggestions everyone!
Duplicate detection is a difficult problem to solve, performance-wise. Constantly checking every new candidate against every existing candidate for a First Name/Last Name match is pretty taxing on our systems, but I agree it would be very useful.
Something like a few of you mentioned, some sort of duplicate report, could be doable however! That way we could have it run in the background and notify you every day or so of any potential duplicates.
We’ll definitely take this feedback into consideration! Always appreciate hearing how our customers would envision a feature like this, it gets us a nice starting point at the very least.
We’ll be in touch!
This has become a bit of an issue for us now that we’ve implemented pocket recruiter. I think a possible duplicate report would save a lot of time and energy when sourcing in pocket recruiter, and occasionally on Linkedin when a candidate profile is not identified by the extension.
Thank you for the feedback. This is something that has been bubbling up as an important feature request so I am escalating it.
Carina: Could you please describe to us in a bit more detail how you are using Pocket Recruiter with CATS.
There has been an internal discussion in the product development team about making our duplicate detection a lot more flexible and robust beyond what we currently have:
Here’s a possible approach we could take. When a new candidate is added.
The system will:
- Match email.
- Match Phone number
- Match full name (last, middle, first)
- Match candidate’s address
- Match resume contents (A candidate’s resume with minor changes will show up as “identical”) since we will be using a “Document Similarity Search” as opposed to a precise search.
The above could be a strict match or a fuzzy match. We could use a heuristic algorithm to create a confidence score (a higher score closer to 100 would mean an perfect match most likely based on email).
Many ATS systems usually only strictly use an email address as the basic criteria for duplicate detection. This works on resume imports and manual data entry but fails when importing a Linked In profile or Indeed or scenarios where only a candidate name and resume contents are available so a more advanced approach described above should solve the problem.
I have asked my team to make this a high priority so please feel free to describe in a bit more detail what kinds of problems you are experiencing with duplicate detection or lack thereof.
Any more thoughts/news on adding the ability for users to run duplicate reports so we can manually fix/check duplicate records? We could really use it about now!
Unfortunately we’re pretty busy here getting ready for our upcoming release. This is definitely something we want to address at the beginning of next year, however. Apologies for the delay, we’ll make sure to keep this thread updated going forward!
I wonder if LinkedIn address would make sense as a criteria for dupe detection? Work emails change with a new job. Personal emails mostly stay same. But a persons LI address never changes and usually has the most up to date employment info anyways.
I second that! If we are building a list of possible duplicates, LI address, email, and names should all be used for matching, then we decide if we want to merge.
Just checking in again - has there been any more progress made on adding the ability for users to run a duplicate report so we can manually fix/check duplicate records? We desperately need it. Even if it just generated a list of people with the same first/last name, I’d be willing to go into each profile to figure out if they’re the same person at this point…
A “duplicate detector” report would be awesome. Just want to chime in that this is something we need and would use! - Amy
Any movement here? This is a great idea that we would use to keep our database cleaner.