Data Output Verification or Mail Confirmation

Options
What do you do so that you can trust the information RE is putting out?


I'm talking about outputs, all the different outputs, queries/exports, mail, all of it. We keep finding ourselves being hit with incorrect data. The query we have for the golf tournament didn't include a merchandise donation. The conditional mail merge didn't produce a letter for the president of the company who is a primary contact on the organization that gave a gift. These kinds of things.


Previous to RE we had some filemaker systems, a custome build Access based database, and the perenial excel sheets. Yet with these solutions we never really had an issue with trusting our data, we had methods to verify.


We're a small office (4) and so we're trying to find efficient solutions.


Here's an example. We had someone exempt from our online donor wall, they called us and let us know (and were justifiably upset) in the constituent query we used to create the donor wall list our criteria was:

Donated between 1/1/2014 and 31/12/2014

Key Indicator equals Individual

Gift Type equals Cash

and Gift Receipt does not equals do not receipted


Logically this should have pulled in everyone that met that criteria.

After phoning BB and 30 minutes of digging around we discovered that this donor had also given in 2014 a gift that was marked as Do Not Receipt (it was collections from a fundraiser and not their actual money.) Because they had ONE gift that was Do Not Receipt they were removed despite their other gifts that were to be receipted.


Looking at it the logic calculation made sense but we changed it to 'Gift Receipt equals one of Receipted, Not Yet Receipted' and then they were on the list. 


I gave this example because it demonstrates how we need to come at the data from several different angles to be able to verify that we have built the query or connected the relationship correctly in order to prevent these mishaps. However it is proving difficult to find efficient solutions especially in cases like with Mail where the output is created before you can verify it against anything.

So here's the question what methods do you use within RE to verify the output? What do your organizations do?

Comments

  • Instead of using MAIL you could use Export.  Then you have better control of the output fields.  You can include more information to make sure you are capturing what you really need.  Then you can record it back in RE using the Import function.
  • Export is a good option, but another is to run several small queries and then merge them together after verifying that your query is working correctly.  Once you merge them together, then you should be able to use the mail function based on your final merge operation.
  • Keri Eadie:

    Instead of using MAIL you could use Export.  Then you have better control of the output fields.  You can include more information to make sure you are capturing what you really need.  Then you can record it back in RE using the Import function.

    We were doing this initially but the conditional merge that lives in mail is a lifesaver. We have 30 different letters we are commonly sending out so it's hard to replicate the ease of Mail. Really the issue is we're looking for some methods to create a second list that we're able to compare the two. It's okay if we're comparing against letters as our batches of letters are typically under 50 so this is possible but we're trying to find good methods to create two lists (or maybe a completely different method) so we can compare to confirm the data.

  • every org I have been at has compiled a list of about 5-10 key names of records that should on a regular basis hit on at least two of the criteria in any given query (whether it be constit codes, give history, volunteer involvement etc.).  When we pull any sort of lists/reports, we look at those names to see if they show up or not.  And if not why.  A way to spot check.  it has worked pretty well.  If you KNOW there are certain ones, like the president, that is not populating, go through the query criteria and compare every step of it to the record that is missing and see where the hole might be.  It could be something as simple as the Send Mail is not checked.

Categories