A Data Governance’s Purpose
I recently read a remarkable book of one endearing dog’s search for his purpose over the course of several lives. The book is called A Dog’s Purpose and it touches on the universal quest for an answer to life’s most basic question: Why are we here? During his countless adventures, the dog, Bailey, joyously discovers how to be a good dog before he passes away. Afterwards though he is reincarnated over and over again spanning the course of five decades. Over the course of his reincarnations, Bailey grows an unbreakable bond with a boy named Ethan. As the boy grows older and comes to a crossroad, Bailey once again comes back into his life to remind him of his true self. By the end of the book, Bailey discovers that his purpose is to teach humans to laugh and love. As I finished this book and started to write this blog on data governance, it made me wonder, what’s a data governance’s purpose?
Enterprises nowadays generate an enormous amount of data during any given product or service life cycle. It gets messy when you have to make the critical decision of which data types to add to your data lake. These datasets can include customer data, vendor data, product data, and asset information just to name a few.
Typically, organizations are only focused on dumping data into a data platform and take little consideration into how it will be used. Also, many times there is no system in place to gauge the quality of data or measure how it has grown—either organically or through other acquisitions. When faced with these challenges, it becomes apparent that a data governance function is needed to better manage how an organization’s information is stored and used. In the remainder of this blog, we will discuss specific data governance challenges and ways to overcome them.
Meta Data Catalog and Lineage
Meta data catalog and lineage is one of the key pillars for data governance, especially in organizations dealing with financial risk management. Some typical concerns that financial regulators want to understand is the source of risk reports. Many times, this data could reside in a data warehouse and ultimately from a set of back-end applications. However, without a meta data catalog and lineage, it becomes incredibly difficult to demonstrate where a specific dataset is sourced from.
In the case of healthcare, employees are required to ensure consistent naming conventions for clinical data. For example, Electronic Medical Records (EMR) could use different clinical coding like SNOMED or LOINC but ensuring some standardization is critical to the quality and safety of health care delivery and is fundamental to the success of eHealth. When diseases, states, drugs and allergies are different for each individual, standards are required to semantically harmonize clinical information into a common taxonomy of medical terminology that can be effectively shared across departments, clinicians and applications.
Based on my experience, the best way to validate the business case for this type of data governance program is to establish a secured, consistent and complete meta data platform with the inventory of all the business and technical meta data. This will almost always surface tangible proof that there are issues related to data and forms. It will also help establish the organization’s group of employees that should come together to define data definitions and rules and discover true data owners.
Improving the Quality of Key Elements
Lack of trust in data quality and lack of confidence in the analytics produced by a data platform can quickly dampened enthusiasm for further investment in data initiatives. To avoid this, it is very important to ensure that you have data quality management as one of your key architectural components.
For example, in financial risk management, risk scores are complex models that calculate the overall probability of a default associated with credit exposure such as a loan or credit card. Because this depends on the quality of various data attributes, the overall accuracy of the score is only as good as its underlying data. This means that the financial risk department should be easily convinced on the importance of owing the data governance plan, while marketing could at the same time benefit from being able to promote the company’s credit score accuracy.
Ensure compliance with regulations
Security and privacy are perhaps the most important drivers for a data governance program. In industries where regulations govern relationships or trade with certain customers, especially in finance, an effective ongoing data governance program could be the difference between compliance and huge fines and brand erosion. Compliance must be an ongoing focus as new regulations continue to evolve in regions all around the world.
It is critical for organizations who are required to meet these compliance standards to have data governance policies in place to discover sensitive information and mask data to prevent unauthorized use.
A Data governance purpose could be described as the formalization of accountability over the management for data and data related assets. The enforcement and formalization requires getting the right people involved at right time, in the right way, and using the right data, and the right platform that will address the right business decision. If a data governance plan is not established then data will be transformed into different variations by various business units. This will also lead to different data rules and definitions, thus lowering the value of the data flow chain, and making it close to impossible for an organization to find its data’s true purpose.
If you would like to define your organization’s data governance purpose, Intersys would be glad to guide you through your data governance journey.