How Data Management Relates to Big Data Success (part 2)
How “Pragmatic” Data Management Works
Whether your big data management plan for success contains two steps or twenty steps, the initial point for most successful companies is data governance. Typically, the focus of data governance tends to be on only inbound data feeds with little regard for the metadata, but if attention is not paid to determining the true purpose of the data, to applying and following business rules, and to managing change, downstream applications will most likely continue to receive lousy data.
Data governance for big data can become complicated, so maintaining practicality is critical. Without a pragmatic governance approach, the value of data is easily undermined. Attaining a high degree of data accuracy is not always necessary because the art of pragmatic data governance is to focus on the purpose of the information and how “fit” the data is for use. That said, if monitoring in-flight maintenance information and safety is the desired outcome, we definitely recommend that the data be very accurate.
The best structure for an individual organization’s data management depends on what the organization is looking for. What is the company hoping to get from its big data? Curiously, a very common answer tends to be “more analytics.” But what is the company trying to accomplish with “more analytics”? Using analytics to drive sales and market expansion and using analytics to minimize fraud and loss both require data management and data governance; however, while the techniques for data management are almost universal for each situation, the focus of data management will differ.
Data belongs to the company, but it is generated by individuals, and someone must be responsible for the data. The organization’s data governance is only as good as its data ownership. Once the correct data owners are identified, which can be a difficult task, especially if they don’t want to be identified, the data owners will need to establish timelines and goals and to make sure that these standards are well-known.
Knowing who the key data users are is also important. As the data moves through the organization’s systems, the users change, and understanding how they work with the data ensures that the governance rules put in place meet their needs. Doing so helps prevent rogue databases and siloed entities that then throw doubt into the process when everyone is supposed to have the same answer to the same question. Once the right people are identified and the skills of the organization are matched to the task at hand, the process of unlocking the value of data can begin.
Communication, education, and training are key to the appropriate use of big data. The inclusion of unstructured data and metadata management is also important. Unstructured data has always existed, but like big data, good ways for capture and analysis were previously not available. Metadata has also always existed, but being able to capture it and use the capabilities to snapshot specific time periods, pre- and post-change is relatively new.
By John Siegman and Jeff Klagenberg, LumenData’s MDM Experts