A Subject Matter Expert (SME) can exist in a variety of sizes, sounds, colors, shapes and formats. It can be a human or a robot. It can even be data that is collected, stored, analyzed, included in formulae and archived. In contrast with a human SME, data is very objective while any subjective angle must come from analysis and interpretation. Regardless of data’s inability to speak audible words, it is able to tell a story limited only by the mind behind the eyes that takes the time to view its rich and meaningful story.
Data is like a box of chocolates, you never know what you are going to get
Data is messy. It can be neatly organized and clean, living in a well-structured Master Data Management system or data warehouse, or alternatively, it could be a duplicative, outdated, and inaccurate blob of a mess scattered around in an inconceivable number of legacy databases that are disconnected. Yep, even with a prideful declaration of confidence that all is well by those who maintain the data, it’s usually evident within a moment of receiving the answers to some simple questions that a “single source of truth” is a foreign concept to many that own the data.
Is the quality of the data really that important? Is the way data is treated and managed by entities significant in the overall scheme of business? Of course, it is! Messy data leads to messy results. If one is relying on the data to provide the basis of quality, reliable results that tell a truthful and accurate story, then the input needs to be quality data. Data science and the tools of today’s technology world are amazing and can clean-up, polish, un-duplicate, associate, stratify, sort and assimilate bad data, but this takes time, effort, and brain power.
So, it’s important to understand the law of the land before relying heavily on the data at play. A considerable amount of work is needed to obtain usable data and to get your SME in shape. On the other hand, many organizations have taken the effort to centralize their thoughts, treatment, and application of data. In other words, they want a credible SME to contribute to the health and wealth of the organization. They desire a single source of truth and workflow processes that leverage data that has already been entered into the ecosystem to minimize rework and eliminate the risk of duplication or dual entry of what is supposed to be differing or same data.
Stupid is as stupid does
Good data is the result of a strategy built on good data practices. The strategy has to encapsulate a ‘Future State’ roadmap to get to the utopian state where your data SME produces reliable analytics and reporting. How do you ensure that the strategy is worth the time and effort?
Before a future state is crafted, consider the current state that data purposes, needs, and uses. This helps address hot spots and, better yet, sheds light on future state opportunities. So, take time to assess the current state of your data before jumping to the future state.
Next, the future state must be a significantly better SME. Given this expectation, some serious forethought must be considered in the future state conceptualization process, including adding business owners, executive leadership, and at times third-party users of the SME into the visionary process. This should be done without fussing over the technology at play in order to stay focused on the data SME and not what helps the SME do its job.
Lastly, it is important to not take shortcuts or set aside important architectural considerations. For example, if Master Data Management, a best practice for centricity of data, is not possible, find an alternative that does not compromise the quality of the future state build.
Run Forrest Run
Getting the best out of your data SME does not happen floatin’ around accidental-like on a breeze. Just like a human SME, you need to persevere with a plan. Because the good thing is that, if you can get the data up and running right, it will be the best SME in your organization.