As much as organizations would like perfect data under their complete governance/management rules and guidelines, those ideal outcomes disappeared already with the big data era's reality. In the "monster data" era, it will be much rarer to have complete data management, especially with the inclusion of more data types, more complexity, and less reliable data sources coming at organizations at an accelerated speed. This writing aims to give some advice on deciding which data sources are likely to get more intense data management. Prioritizing will allow data management to focus better on balancing its efforts. It will also enable the data usage to be separated from Ideal/excellent data management allowing new data resources to enter the organization and grow into data management over time. Prioritizing data sources for robust data management should consider goals, measures, and change channels.
Organizations can't focus on all the data coming their way, and they certainly can't depend on AI supercharged data scientists to find interesting anomalies from the bottom up. The first cut at the data under management should directly relate to business/organizational outcomes necessary as a base. Many organizations set up strategic goals that represent the most desirable for the circumstances at hand. Savvy organizations plan alternative strategies/scenarios and practice the most likely emergent strategies. Most organizations, these days, tie both their functional and operational goals to the strategy du jour. Savvy organizations look at the end to end processes/journeys to link goals together to minimize organizational conflict and consider constituent outcomes as legitimate goals. These efforts would include customer journeys laced with highly desirable customer experiences. These goals will help narrow down the data sources that need proper data management, although much more than in the past.
Prioritize on Real World Measures
An excellent place to start is where most of the data activity occurs. Many organizations create heat-maps of highest use and apply the highest level of governance for these sources regardless of whether they are operational or decision focused data. There more enlightened organizations push the data timing to real-time, representing the Database of Now. These same organizations practice data mining to look at the reality of the current and past situations that may point to future operation optimizations. The aware organization also mine new data sources like voice and video for customer experience improvement potential. While much of the mining is descriptive, the more advanced efforts start prescribing potential for change and even predicting future behaviors. In this kind of environment, decision optimization is considered a major important improvement arena.
Prioritize on Change Channels
Like it or not, things change and almost always get more complicated. Rapid change may mean cobbling together creative responses in a quick fashion. To that end, organizations need to sniff many data sources to change potential impacts that could affect strategy or operations. Sometimes it might be subtle to look at signals and events in these data streams, but it could range to finding emerging patterns within or across data sources. The likely channels would often include competitors, markets, industries, demographics, vendors, suppliers, geographies, legal frameworks, and governmental pushes. Additionally, best practices involve finding new contexts and sources.
Figure 1 Optimal Data Governance Defined
Applying excellent data management governance, depicted in Figure 1, is no longer possible to all data resources. Organizations will have to focus on the data sources that are key for business outcomes that enable organizations to survive, thrive, and even optimize their impact. While Data Management is necessary for specific resources at a high level, the time has come to identify those particular resources and set data governance levels across various data sources.