When we went live with Tessitura in 2018, we imported a lot of data that had not been maintained over the years. I spent a lot of time in 2020 going through records of people who are actively on our mailing list to look for deceased (some had been gone 10+years!) or evidence that the address we had for them was no longer valid. But there are a lot of other records that are just sitting there because they aren't on the mailing list and they haven't purchased a ticket or made a contribution since we went live. As we come across these kinds of records we have been inactivating them with a reason of "No Activity." But I'm just curious if other organizations have policies surrounding when to deactivate records. And does it matter if they are active if we aren't pulling them into lists? We run our mailing lists through NCOA via our mail house (and I update changes), but we aren't currently using one of the integrations for NCOA. On the one hand, they are just sitting there and not being pulled in, but I hate the idea of a bunch of inaccurate and unused records just sitting there. Interested to hear how others handle things.
Anne Robichaux (she/her)CRM AdministratorThe Historic New Orleans Collection
For us, unless the person is deceased we keep the record active.
There have been a few times, especially recently, where a patron has come in wanting to renew their membership. When we looked up their constituent record, they hadn't renewed or donated or bought tickets to anything in over 10 years (sometimes longer). Our Front Desk staff doesn't have access and wouldn't know how to reactivate one of those accounts had we made it inactive, so to avoid duplicate records being made...we keep everything!
I'm currently in the lengthy process of cleaning up our records (which hasn't been done since we switched to our old system almost 10 years ago). Almost 100,000 constituent records is no joke!
About 10 years ago now I had the box office staff inactivate a list of constituents that did not have recent history with "no activity" and I regret it. We are now getting those people returning and I am reactivating their records and merging with the new records they have created. My recommendation is to just let them sit. They aren't hurting anything being there. Unless of course they are deceased.
We don't even inactivate deceased records, as this often causes problems for staff who need to finalize adjustments, record estate transactions, managed "In Memory Of" soft credit, etc. We make sure to pay close attention to the "Deceased" status on customers for any communication List/Extraction, and that means making sure that a household with all deceased members is also flagged as deceased, but it's definitely worth it to us to not have to do "active/inactive" management, or worry that someone's information is not showing up in a List/Report/Constituent Search because of their inactive status.
Thanks, y'all. You've convinced me. I'm going to keep inactivating some of these really old records that literally only have a first and last name, but I'm going to do more cleanup on the others. Some of these records were created in the early 90s, so the chances of the data being accurate is not good.
We also make deceased patrons inactive (and of course we do so as well with bot/spam accounts; that is a given). We have seen no issues arise through our processes due to this, and in fact our Development department appreciates it.
The BIG thing I think, is to make sure your patron activity matches those accounts you are pulling for mailings/e-mails. Yes, the patron who bought tickets 11 years ago might still live there and still use that e-mail address and still might WANT to buy tickets to that one event that is almost exactly like the one they attended 11 years ago. But is it really worth still including that patron in your lists? Certainly you COULD make it inactive to make sure it steers clear. But I would just have better criteria in my extractions.
John A. Moskal II said:We also make deceased patrons inactive
Something we've tried to do (in the before time at least) is to convert deceased constituents that we expect to do some/any further activity with to customer type = Estate, which leaves things active, transactable, etc., and keeps them out of marketing pulls.
A discussion that has gone back and forth internally. Our Development Strategy person (that team's data person and my #1 counterpart in that department) is of the opinion that such things should be given separate accounts and treated as a Foundation/Estate based on the circumstances because in that case it is no longer that person transacting but a separate individual or group of individuals transacting in the name of that person.
I can see both sides to the argument here, but since she is so tight in her own data practices (I love not having to come down on Development for tidy data because she does it for me), I figured I would let her have her way on this one. As long as each living individual only has one account, I am happy (I had to win that fight about 5 years ago where multiple accounts would be created under one person's name if they were doing things to be mailed to work or home). So I am good enough with this approach.
The good thing with this particular subset of old records is that nothing is being sent to them. They don't have any activity since we moved to Tessitura so they aren't getting pulled into extractions. Our general mailing list requires a contact point purpose on the mailing address and there really has to be something to trigger getting pulled into an email list (other than just having an email list). We didn't have a ticketing system before Tessitura, so any historical data lives on the Promotions tab (and we don't use that for lists).
We have put a very clear set of criteria in place for actually purging the records using the Purge utility. With GDPR in the UK we have to prove that we are only keeping records for as long as we have a valid reason to keep the data. So we would only keep them whilst they are still actively interacting with us within a period time we have defined, based on analysis we did on how often people return to us.LIke you Anne we imported a lot of data and that hasn't been touched in years and so we want to get rid of any personal data that we don't need. The postal code and locale will still exist along with the ticket history, so we can still use the data for analysis, so we're comfortable with purging the data.
Caryl
Hi Anne, like Caryl we have a set of purge criteria. I run it weekly in Segmentation/Extraction Manager. We too went live in 2019 with a load of data that had never been cleaned before and come from two previous ticketing systems, from a time when the team (30 years ago) didn't even take addresses or even first names etc, it was a mess!. Our first purge using the criteria set was for 38k records if I remember rightly. We got support to do those in bulk for us as it was too many to do manually and not cost effective. It was about a fifth of our customer records!! It is visibly noticeable when you are working in orders with the amount of constituents returned on search. It also helps with the merge tool returns. Our criteria looks at criteria like e-newsletter subscription, and also if they have even opened an email that has been sent to them. We use 6 years as our cut off so that any HMRC tax queries regarding gift aid can be easily answered - I don't think you have that in the US though. We didn't want to get rid of anyone who was still opening our emails, even if they hadn't bought a ticket, but that might not be everyone's choice. We also don't purge any business or school records either - that's all done manually by the respective departments. I'm happy to answer any questions, feel free to msg me. Each week we have about 20-60 records that come up in the purge count that need setting for purging...its really worth doing.