What is worse than bad data? Bad data without a strategy to make it better to best data with constant cleanup. What’s worse yet, rolling out your strategy and having missed critical steps that then wreak havoc on your data customers (SFDC users, etc.). Devising a data quality strategy is not just about cleaning bad data and getting it done. There are many areas that are overlooked and/or undervalued. Here are those key areas to consider in your data quality strategy plan:
1- Defining Root Causes- Conduct data input audits. Review forms, uploading rules, integration input standardizations, field standardizations (telephone, state/province, country), if there rules for automating the correct standards for data entry, and email validation processes.
2- Cleaning Impact- When cleaning existing data, it is imperative to reduce the risk of triggering off workflows, emails, etc. when cleaning data. Define priorities for batch cleansing with an organized plan for turning down alerts across systems to avoid unnecessary interruptions to your internal data customers. Word to the wise, weekend cleanups are the way to go for any data set to enable your ability to pull back and roll back data.
3- Backup Data- Before you do ANYTHING, be sure to take full back ups of your data. No questions asked on this one.
4- Communications With Your Data Customers- Make sure you have a communication plan for those who are going to be impacted by data cleanup and standardization efforts. Send communications a few weeks before things get started and be sure to engage leadership in the data customer communities to be a part of those communications. There should be ongoing communications once launched and for any ongoing maintenance runs. Ignorance is NOT bliss when it comes to changing data on data customers!!!!
Also, although this is a little stale, I still think that this covers a lot of great information with links to great resources on data quality: http://topliners.eloqua.com/docs/DOC-2402
Until next time!