Poor data governance often leads to a failed data state. Too many times companies, and even silos within a company, ignore the importance of data governance. Such neglect leads to increased cost, decreased compliance, and even a complete failure of a data solution.
A key data governance principal to adhere to is data usage. This is a simple one. If your data is not being consumed, then it shouldn’t be stored. Period. From time to time in our industry we come across a department/company that we think should be on A&E’s and TLC’s Hoarders television shows. In all seriousness, the case can be made that data hoarding is more destructive to it’s environment than those possessions taking place in those television shows.
What is an example of how unused data can be destructive? Project 1-A wanted to store person name and address in a single database table so the company could easily send out marketing information to the customers for a new product. The information that project 1-A needed was already in the database in multiple tables, but it was too complicated so they created their own table. Project 1-A was in the process of using the list to complete their mailing when an architect stepped in and informed them they needed to use the existing data which is kept up to date with address changes, name changes, do-not-contact lists, etc. Close call! After project 1-A is complete, the newly created table stayed put for years. Project 1-B comes along and needs a mailing list to send more marketing materials to ASAP. Project 1-B finds a table in the database that contains names and addresses together. Awesome! They decide to save time/money and use that list. They send out a million mailers at a cost of $400,000. Unfortunately, the list was out of date and 25% of the mailers were returned and another 10% of the mailers were sent to do-not-contact addresses. Do-not-contact lawsuits are pending. Yikes!
Some would be quick to place the blame on the project for using the data, when in reality, the data should have never been there for them to use in the first place. If the company had a data governance process in place that identified unused data, the data would have not been there for someone to use unintentionally. A simple data usage report in Tableau or similar BI tool would have informed the company that the bad data has not been changed and/or accessed in years.
Remember, bad data is NEVER better than no data! Don’t hoard data just in case you may need it at some point. There is quite a bit of cost savings with trimming the fat in your data warehouses as well as avoiding problems like the example above. Put a process in place to identify your most and least used tables and databases. Create a procedure to sunset those unused tables and databases. Your company and your reporting teams will be much better off!
Happy mining,
-Eric Mayberry
To read origina post, click Here
