Database Clean-up Project (Database Health Tools)

Our organization has never used any of the data health center tools and I am about to start trying to run some of them for the first time ever. Our organization has been around since 1976 and definitely needs a good clean-up/overhaul.

Has anyone ever used: validate addresses, address finder, deceased finder, email finder, phone finder, age finder etc? It could be either regularly, or as an initial first time use project like I'm planning.

We just wanted to know if there was any tips or tricks or anything we should be aware of orthings we should look for before we hit go on utilizing any of the tools.

Thanks!

Comments

  • @Jordan Carta

    Definitely work with Altru to provide a specialist to help you walk through the process the first time. There are a lot of nuisances to the various settings that are somewhat well documented, but they don't make a lot of sense until you've done it at least once.

    Once you commit a batch, you can't undo it. It's VERY easy to drastically change a lot of data at once.

    The different tune-up steps allow you to review the data that's going to be changed. It can be a bit overwhelming, so give yourself lots of time.

    We've only ran deceasedFinder and addressFinder as the other dataTuneups are a bit pricey. I would run deceasedFinder first to cull out records that don't need to be updated, then work on the other tune-ups. I'd also probably stagger them, so you're doing maybe one tune-up a month, if not further apart, so you can better assess any problems that arise from the batch update.

    My other tip would be to absolutely ensure the “source" for changes is marked as dataTune-up (Address for addressfinder, etc.), so you can better track issues.

  • @Jordan Carta did you end up using those tools? I've gone thru the duplicates for my museum and now seeing what other tools/steps to take in the data clean up process.

  • @Jordan Carta, AddressFinder works pretty well, although you can only submit addresses in the US and only Address1 less than 50 characters.

    DeceasedFinder requires careful review, since it doesn't take into account generational suffixes (Jr, Sr, etc.) and since it tends to have false positives on common names.

    For duplicate management, I recommend setting up a tiered approach (99%, 95%, 93%…) to make it easier to deal with low-hanging fruit first (i.e., true duplicates) and the more complicated ones later (those that require research or to have households created). Also, create your own merge configuration so that you merge everything, including duplicate contact info.

    If you need help, please feel free to reach out: knight-consultants.com

  • @Andrea Fisher

    Yes, we have started using some of the tools. I turned on address accelerator towards the beginning of the year for any new constituent records just to see how it worked. We gave that a few months to see what everyone thought and no one had any complaints so my supervisors gave me permission in august to actually run it against the entire database to fix all the incorrect formatting etc.

    We had a few weird things pop up because of internal addresses and c/o addresses but we were able to use the query it generated to filter by anything “weird” so we could fix addresses as necessary. Once we ran that we ran address finder and haven't noticed any significant issues since running it.

    After running both of those we ran deceased finder and are manually working our way through the list that was generated in order to verify and/or to find if any spouses of the deceased are also deceased or not. For the most part we haven't found many issues other than a couple of Jr/Sr situations and a few individuals who we received gifts from after their death date but most of those have either had a surviving spouse and/or were from one of our imports from an outside gift source (ex: we are a CMN partner and some of them were collected by CMN).

    We are being extra careful because we have a lot of older records which go beyond the range of the address finder NCOA date range, but so far we honestly haven't come across many incorrect matches from deceased finder.

    If you decide to try any of these tools I would definitely make a plan for how your organization plans to deal with the results. we decided to keep exports of everything so we could reference back to any of the changes just in case we needed to, but your organization may not want to do that. We also did it one step at a time to “test the waters” before we tried something else as we weren't really sure what to expect but if you're more comfortable with doing them all at once I would definitely go in the address accelerator, address finder, and then deceased finder order so that if any addresses were changed before someone passed it is reflected in your data so that deceased finder can match it to death certificates whenever possible.

    You should also decide whether you want to set up the tools to run automatically on a schedule or if you always want to run them manually (ex: address finder can be scheduled to run automatically quarterly). I would also recommend running deceased finder at a time when your organization has a slightly less busy schedule in general so that you have time to sort through the returned matches if you are going to do any verification as it is a slow process.