Deleting records with no activity

Agata Ketterick 3

I'm new in my role as a database manager in a higher ed setting. I've noticed that our database has a lot of constituent records that don't seem to be in the database for any good reason (i.e. individuals with no address, no gifts, no activity; local organizations with no activity). On the one hand, why not just leave them in the database, but on the other hand, these records often pop up when I enter new constituents as potential duplicates (and maybe they are duplicates, but it's impossible to tell since the records have so little information to begin). I hesistate at just deleting them since maybe there was a good reason at one point to put them in, but I really can't see the reason and no one at my organization can think of a good reason either. So I'd like to get this community's opinion on what I should do! 

JoAnn Strommen

As I'm a fanatic about maintaining a clean database, my vote is delete based on your description: no gifts, no address, no activity, unknown if duplicates, etc.  You didn't mention - do they have relationships?   Does looking at the date added give you any clue as to why they may have been entered?  Perhaps following an event. 

I know other users may disagree but I would delete them based on what you've described.  If you can group them you could export them to Excel just for historical purposes but I don't see any reason to do even that. 


Just one opinion...

Jen Claudy

There are a number of threads in the Community about deleting records.  In your case, if the only data on the records is a name, then I would be inclined to delete them.  For local organizations, you might keep those if you think you'll ever make an ask of any kind...and you can look up their address info on the internet to complete the basics of the record.


You may want to look at the Constituent IDs and also the Date Added and Added By fields to see if these were all added at the same time and/or by the same person.  Might shed a little light on how/why they were created originally.  Those fields are available in Query or if you spot check, the Date Added and Added By information is in the Properties Window (icon of a hand holding an index card).  Import ID may also prove useful...when I Import a set of records, I assign the Import ID and include something that tells me what Import these records were part of (for example, we have a program that goes by the acronym HCC internally, so records I import from that program have Import IDs that start with hcc-).

Agata Ketterick 3

A large number of these records were entered in 1997 because they were invited to an event. If they had a mailing address it would be one thing, but without a mailing address, they are driving me crazy!

JoAnn Strommen

19 year old records with no info - go for it.  They would bother me too.

Ryan Hyde

Agata Ketterick:

A large number of these records were entered in 1997 because they were invited to an event. If they had a mailing address it would be one thing, but without a mailing address, they are driving me crazy!

Yikes, yes, get ready to purge those records! But I do mean "get ready," as in make sure your queries are robust and maticulous (scour records for usable data and don't include those in your first round purge). Then output some data in your query and just start dipping in and seeing what the records look like to make sure you're not overlooking some type of data that is actually useful. You just want to make extra sure you're pulling the trigger on the exact correct subset of records.


And to answer your initial quesiton of "why not just keep them?," the answer is that someday we're all moving to NXT, and whenever that happens, you're going to be migrated to a per-constituent-record pricing model. So these useless records will eventually cost you money. Best to think about doing this now, when there's not so much pressure.

Jennifer McHale

You can always request a back-up of your database, export records, etc.

As someone who is now frantically cleaning up our database to transition to NXT ... I'm all for deleting! I'm finding a lot of names were added, but not a lot of information or, certainly, no way to ever pull the data is a meaningful way. I believe many of these records were added for a long-ago mailing too; frankly, I'm not so sure most of those people would even live at the same address this many years later.

So, I have the back-up copy of our database, should we ever need it. But, I'm trying to get us down to the records that are most relevant to development and communications.

Daniel Noga 2

Agata Ketterick:

I'm new in my role as a database manager in a higher ed setting. I've noticed that our database has a lot of constituent records that don't seem to be in the database for any good reason (i.e. individuals with no address, no gifts, no activity; local organizations with no activity). On the one hand, why not just leave them in the database, but on the other hand, these records often pop up when I enter new constituents as potential duplicates (and maybe they are duplicates, but it's impossible to tell since the records have so little information to begin). I hesistate at just deleting them since maybe there was a good reason at one point to put them in, but I really can't see the reason and no one at my organization can think of a good reason either. So I'd like to get this community's opinion on what I should do! 

My org is migrating to Altru next year, which, like NXT, is hosted in the cloud and I'm pretty sure you pay per constituent record with Altru too. I'm in a very similar boat. Most of the "duds" that show up in my query do have addresses, but as others have mentioned, they are old enough that I can't be sure those people live at the address anymore. Some were people we invited to our groundbreaking when the museum first opened, but who never responded to us. With no gifts, no event attendances on record and for lack of certain Attributes we use to flag records for solicitations, there's no way they'll ever appear on a solicitation list that we pull; we'll never contact them again, and if they ever contact us, it will be of their own volition (in which case it's not much trouble to add them back to the database).


If a record has no immediate relevance to you or anyone else at your organization, it's dead weight. Only in a very few cases I can think of is "some day, maybe" a good enough reason to keep them.




 

Daniel Noga 2

Ryan Hyde:

Agata Ketterick:

A large number of these records were entered in 1997 because they were invited to an event. If they had a mailing address it would be one thing, but without a mailing address, they are driving me crazy!

Yikes, yes, get ready to purge those records! But I do mean "get ready," as in make sure your queries are robust and maticulous (scour records for usable data and don't include those in your first round purge). Then output some data in your query and just start dipping in and seeing what the records look like to make sure you're not overlooking some type of data that is actually useful. You just want to make extra sure you're pulling the trigger on the exact correct subset of records.


And to answer your initial quesiton of "why not just keep them?," the answer is that someday we're all moving to NXT, and whenever that happens, you're going to be migrated to a per-constituent-record pricing model. So these useless records will eventually cost you money. Best to think about doing this now, when there's not so much pressure.

 

I second these thoughts; I've got about 7,000 records to review manually in a Query List. Most of them will end up getting deleted anyhow, but there are so many stray, undocumented Attributes in our system from days gone by that I often need to make this judgment call on a case-by-case basis. Sometimes the mere presence of coding isn't enough to make that call; context can be important too and no query can account for that.

Amy Barker 2

If they are very very old and have no other information than a name, I would delete them. If the only activity is on the Bio I tab, and there's nothing else on any tab, and it is only a first and last name, they're just clutter.

Rachel Slager 2

You work in higher ed, adding to the great list of other's above, just make sure that there are no usable phone numbers or emails or education records.  We are obligated as universities to report alumni who we have some form of contact for even if they don't give and don't have an address.

Christine Cooke 2

Jen Claudy:

There are a number of threads in the Community about deleting records.  In your case, if the only data on the records is a name, then I would be inclined to delete them.  For local organizations, you might keep those if you think you'll ever make an ask of any kind...and you can look up their address info on the internet to complete the basics of the record.


You may want to look at the Constituent IDs and also the Date Added and Added By fields to see if these were all added at the same time and/or by the same person.  Might shed a little light on how/why they were created originally.  Those fields are available in Query or if you spot check, the Date Added and Added By information is in the Properties Window (icon of a hand holding an index card).  Import ID may also prove useful...when I Import a set of records, I assign the Import ID and include something that tells me what Import these records were part of (for example, we have a program that goes by the acronym HCC internally, so records I import from that program have Import IDs that start with hcc-).

In addition to what Jen Claudy has said -- Keep in mind you said you are at a higher ed org.  That means you have alums that will never necessarily have any activity on their record, but you still have to keep them!  Those orgs that seem for no good reason may actually be somehow related to someone in the database and someone did not do the housekeeping to link the records.  My first inclination (and I have done it in more than one database) is to pull out a list of the Org records and compare the names -- because I have inherited databases where the orgs were entered mutliple times but named differently (The ABC Company, The\\ABC Company, A.B.C. Company, ABC Company, Adams Birch Congo Company) and I had to go in and merge all the info once it was determined they were the same org.  I cut the list of orgs in about half doing that.  Also, if you are an org or education org with any relationships to the church systems you will find that some records are listed at St. and some as Saint or Saints and those should all be consistent.  AND for all of the examples listed, there should be an alias for each variation so the dupes do not multiply again.

Phillip Bicknell

Regarding migration to NXT and its pricing structure being on records not users, at a recent Blackbaud user day in London UK the sales team indicated "talk to us" about old, rarely used records, implying that cleaning them out of RE7 wasn't a prerequisite.