Uncertain duplicates in NXT - leave them as potential dupes or resolve them all?

Katie Cagney

I have a query on how you handle non-dupes-which-might-be-dupes.


I have 5k duplicates in NXT matching only on name. I'm loath to go through them all and mark them as non-duplicates, partly because it will take forever, and partly as they may be duplicates (I just don't have enough data on them at this stage). Leaving them of course would save me a lot of work.


And yet, I hate to just leave them as my database tidying mind wants there to be NO DUPES!


I am assuming that even if new data is added to a constituent, if they were marked 'not a duplicate' they wouldn't show up in future searches even if the two records now shared, say a duplicate email address, for example, and that got me thinking whether I should be resolving them all.


Anyone got any wisdom? Do you leave dupes you can't resolve or do you process them all as non dupes if you don't have the evidence to say they are the same.


Thank you!

Katie

Miki Martin

Katie, I feel your pain! I think we have about the same amount of duplicates, maybe more, and knowing that drives me nuts. Many of ours are tribute records, though, so that presents its own issue since we don't have the Tribute Module.


As for what to do, when I run across possible dupes, I take care of them immediately. Many of ours are just a name and possibly an address, which helps. I merge them and record an Action, noting the merged record. I only merge fields that have information the target doesn't have, which is usually all contained in Custom Fields for us. I'm not going to merge the summary or contact information because I already have it and don't need more fields populated in that record; that's just even more to clean.


If I cannot say for sure the records are duplicates I leave them. It's better, in my mind, to have two out there. We do have instances of similarly named people in our database that are related, like father and son, that are different. If I come across records that have common names, or that I can determine are not dupes, I try to add additional information to them and also mark them as not dupes.


I know there's the ability to handle duplicates globally but other than deleting those I've marked Inactive via NXT I have not done anything with that. I know it's more time consuming handling them one record at a time, and it keeps many unnecessary dupes in our database longer than I'd like, but I just cannot bring myself to handle them en masse.


Last Tuesday they offered a webinar on cleaning up duplicates. It was helpful, as I did not realize you could do different actions in the database view as opposed to NXT to capture different information. Fortunately, I don't think it will affect our records I've already merged. Here is a link to it. You can also find it in the Knowledgebase under Webinars. Spring-Cleaning Your Database: Strategies for Duplicate Management is the name.


Hope this helps. Good luck!

Dennis Ladnier

Similar to Miki we take care of them immediately hence no large backlog. With that said, it was tough to do the initial catch-up but once done it is smooth sailing.

Dariel Dixon

I don't think it's bad to have a couple of potential dupes out there. If you can't say for certain if they are or are not a duplicate, I think it's better to leave it. You may get more information later to be able to determine if the records are dupes or not. I know it is aggravating to have a bunch of duplicates out there, but if you have removed all those that are true duplicates, I think you're moving in the right direction.

Karen Diener

I think what everyone said so far is very true. The only thing I want to add is to accept the fact that you will never have zero duplicates. We've all inherited databases that suffered from poor data management decisions. A lot of those end up in the category you identified where you'll just never know for sure.


If you do not know for sure that they are duplicates, leave them be and move on!


Karen