Everybody needs their fix … of coffee or tea that is. Your local barista knows your tastes. Personally, I am not a warm beverage fanatic. But I know how I take my data. I’ll take my data big and smart.
What Does that Mean?
Most people have heard the term Big Data. Big Data is coming from more and more places, coming at you quickly and in many different formats. Some people use 3 or 4 V’s to describe Big Data – Volume, Velocity and Variety. Recently I have seen the word Veracity used to describe Big Data as well – meaning accurate and reliable data.
Companies that are harnessing big data thus far report inconsistent levels of success. That fact should not be surprising. Costs are high and returns are not always immediate. Data science is complicated. Improving analytical capabilities requires the design, build and use of complex databases that integrate less-than-perfect data from across the organization. Companies are hiring Chief Data Officers to address Big Data challenges at an expanding rate and many CDOs are just getting their feet wet when it comes to executing strategies and delivering on their roadmaps.
The truth is, many of these companies were less than optimal in how they managed their data before it grew so big. Managing Big Data requires planning, resources and formal governance of that data. I’ve mentioned governance in this column before. Data Governance is the formal execution and enforcement of authority over the management of data and data-related assets. Optimizing the value you get from your data requires that you execute and enforce authority to deliver accurate and timely data for purpose. Governance becomes more important as the data grows. I have experienced this first hand. I help companies put data governance programs in place.
What Makes Data Smart?
Smart Data is harder to nail down. Smart means having or showing intelligence. Making data smart requires attention to detail. The details have to do with how that data is defined, produced and used across the company. The details of the data include information about the data – where it is available, the standard meaning of the data, where it came from and how it can and cannot be used. The data does not become smart on its own and requires that people have formal accountability for the management of that information about the data.
Metadata, or “data about data” as data guys like to say, helps to make data smart. Companies that effectively manage their metadata often manage their regular data properly as well. Metadata management is an entire industry in itself. There are tools that companies use to record and make available information about the data they define, produce and use. Metadata and data governance are like twin sons from different mothers. Data governance results in metadata and the metadata itself must be governed.
Formalized accountability for the management of data also helps to make the data smart. Left to their own accord, or should I say without organizational oversight, different divisions of the same company will typically define, produce, and use their own data. Data ownership is a commonly used term inside many organizations. The “owner” is responsible for deciding what can be done with their data.
Calling these people Data Owners is a misnomer. Nobody in the organization owns their data. The organization owns the data. Allowing people in the company to own their data does not lead to Smart Data. In fact, Data Ownership – the personal possession of a corporate asset, leads to the development of silos of disparate data.
Organizations are moving away from using the term data “owner” and switching to a more appropriate term of data “steward” – defined as someone who is responsible for taking care of something for someone else, in this case data and the organization.
Data stewards play a major role in data governance programs far and wide. Data stewards are managed differently depending on the approach that is taken to govern the data. These people can be specific individuals with specific responsibilities for specific data, or they can be everybody in the organization that is held formally accountable for defining, producing and or using data as part of their daily job. In my non-invasive approach to data governance, I suggest that the latter.
Turning Big Data into Smart Data requires formal data discipline. It requires that somebody is held accountable to Senior Leadership for cross-organization governance of that data. Turning Big Data into Smart Data requires that organizational data is formally stewarded as a valued asset of the organization. How well are you turning your Big Data into Smart Data? Is there someone that has that responsibility? Do they have a plan? You know … it’s all in the data.
Ordering up your regular Cup o’ Joe may be easier then ordering up either big or smart data. Many companies and organizations want their data to be as big and smart as possible. In the future, the company’s success or failure may depend on big and smart data. Big and smart data begins with solid governance and metadata management practices. How do you take your data?
