A new MDM initiative is fraught with many challenges, starting with making sure that people understand what master data management is. Up next is choosing your first master data domain (i.e., product, location, asset, party or one of its roles such as customer). Then, just when you thought the going should be getting easier, it is time to confront another challenge — choosing your master data sources.

“Eenie, meenie, minie, moe” is a children’s counting rhyme used to select a person to be “it” for games such as tag. Prashanta Chandramohan (aka the MDM Geek when his party role is blogger) recently blogged about Identifying the Right Sources of Master Data, which made me think that “eenie, meanie, mindie your MDM sources” would make a great counting rhyme to use for remembering the steps involved in selecting a data source to be “it” for providing master data.

“Getting things straight at the beginning,” Chandramohan explained, “is a critical aspect of the MDM project as it acts as a foundation for future source system integration plans.” He recommends starting by listing the most trusted data sources your organization currently uses, realizing, of course, that this is easier said than done since most organizations have multiple data silos of master data used within a specific line of business, and especially across lines of business.

Data profiling early and often can help with evaluating master data sources (the “eenie” part), but further complicating the data quality challenges (the “meanie” part) of MDM is that each division, department and business unit across the organization will consider itself to have high-quality data from the perspective of its own business users. Which is why the success of your MDM initiative is less about the capabilities of your MDM technology and more about the communication and collaboration (the “mindie” part) among all of the people involved with your MDM implementation.

Playing “eenie, meanie, mindie your MDM sources” might not sound like the best way to spend some of the time on your MDM initiative. But after you say “tag, you’re it!” to the master data sources you will use, your organization will spend a lot more time, effort and money on MDM. And after the first phase of the implementation is rolled out to your business users, you certainly don’t want to hear them say: “M-D-M stands for Master Data Mess, so when it comes to my master data needs, MDM you’re not it!”