One could argue it has become cliché to make references to the enormous significance and proliferation of data globally. It is broadly agreed that the size of the digital universe doubles every two years at minimum. Human and machine generated data is increasing even more rapidly at 10x that of traditional business data . By this coming year, there will be around 40 trillion gigabytes of data assets under management and 97.2% of organizations will be investing in Big Data and Artificial Intelligence .
Consequently, few would argue with the assertion that data governance has never been more vital to business operations than it is today. Why then is it still commonplace to hear of corporate data governance programs still struggling to offer a self-evident value proposition to the business? Why does it remain the case that businesses will temporarily “get religion” for data governance only after something goes terribly wrong, only to return to old habits again in the future? Why, above all, do we continue to hear of failed data governance implementations?
NO IVORY TOWERS
Here’s how you blow up your data governance implementation: set up an Ivory Tower and spend your days pontificating on your next great policy. You may laugh about the absurdity of that statement, but it happens time and time again.
Data governance is not a back-office function. It is the responsibility of a data governance team to connect their program to the business. This is not to say the data governance office should be a function within the front office itself. However, in order to be effective, a close relationship must be established with the front office to ensure the goals of the data governance team are tightly correlated with the business. In fact, the business should feel confident that their goals are your goals too, and your role is to ensure their goals are met without any 11th hour surprises. Trust me, they will appreciate you for this.
The data governance office should also have a tight relationship with the legal team. If done effectively, the data governance team should be heavily involved in discussions involving legal, data, and the business. You will find that when situated properly, the data governance office becomes an important contributor, and indeed a moderator of sorts in these discussions. While the legal team will often err on the side of caution, the business needs to take on certain risk in order to thrive! They will rely on you to help find the common ground that is best suited for the business overall.
You should never produce any policies that haven’t been vetted with all stakeholders. It is your job to ensure policies are well understood, agreed upon, relevant and actionable by all parties involved. Remember this: if there are any 11th hour surprises, it is most likely your fault!
SO, YOU’RE CONNECTED WITH THE FRONT OFFICE – NOW WHAT?
Ok, so you’ve established the needed relationship with the front office. That’s great, but will this be all about policy setting? No, it shouldn’t.
If your product is data, or is derived from data, it is a near guarantee your customers have a keen interest in knowing how well your data is governed. They will be very interested in understanding how confident your company is with respect to its data and ongoing regulatory compliance. Who better to represent that than the data governance team? Speak to your business partners about this.
Your customers will also have valuable feedback about the quality of your product. Again, if your product is data, or is derived from data, this is mission critical information. The customer perception must be understood by the data governance team in order to assess what is working well and what isn’t from the customers’ point of view. Remember, perception is reality. The data governance team should play a large role in understanding the customer perceptions and, when necessary, reversing them.
So, ask yourself this: when was the last time your data governance team spoke directly with customers?
THE GOLDEN RULE
If there’s one thing to always remember about data governance, it’s that people do get it. Where data governance gets a bad reputation is the perception of unnecessary overhead or bureaucracy.
Imagine this not-so-fictitious example. Whenever Johnny in the Customer Service team enters ‘-999’ in that CRM field he doesn’t understand, Jane in the resolver group rips her hair out every time it shows up in her reports. You can believe they both understand the value data governance can bring to their lives.
Data governance is primarily about people and process, and the golden rule of data governance is that people basically want to know what they need to do and when they need to do it. They do understand the value of working in a coordinated and agreed upon fashion, and they know the results would benefit everyone. Where things go awry is when the solution is more of a burden than the problem itself.
While it would be ideal that Johnny only be allowed to enter valid, pre-defined values in the CRM system and that these values should be mastered somewhere independent from the CRM application and Jane’s report, very often there are practical reasons why this doesn’t happen (conflicting priorities, lack of funding, etc.).
Offering that the data governance team will backlog this issue with a long-distance ETA is probably not going to be met with a round of applause. While logging and tracking the fix is advisable, training Johnny and his peers on the proper use of this field and making the CRM data dictionary more understandable for the users would be a likely win for Jane and quite a few others too.
Automation supporting data governance workflows in a manner that is non-disruptive to business users is always advisable. Do not underestimate the importance of the user experience, however. If the tool is difficult to use, you will have difficulty with adoption. Keep it simple and focus on conveying what needs to be done and by when. Do not spam users with emails. Instead, reach out for a discussion when there is a need for attention.
SURE, BUT HOW WILL WE SCALE TO HANDLE EXPONENTIAL GROWTH?
Along with exponential growth in the velocity and volume of data, there will inevitably be an increasing need for data governance at a massive scale. It is anticipated that the volume of machine-created data alone will increase at 50x its current growth rate . In order to keep up, data governance itself will need to rely on Artificial Intelligence (AI) and Machine Learning (ML) in order to sustain itself.
To truly be on the cutting edge of data governance, we must begin to think of a future where Data Governance as a Service (DGaaS) exists.
In the same way AI and ML are used in conjunction with Deep Learning to identify previously unknown relationships, the foundation of DGaaS will be meta-driven AI, ML, Natural Language Processing (NLP), Deep Learning, and Semantic Graphing. Together, these technologies will automate the process of identification, classification, security, privacy, lineage, and indeed even the act of data stewardship itself.
For example, imagine a cell phone carried by a patron at a dance club “listens” to music, identifies the track, sends the title along with the geo-coordinates to a service that identifies the location and determines whether or not the dance club was licensed to play the track. If you think that sounds far-fetched, click on this link . While it is a given that such levels of data governance maturity have yet to be attained, the velocity and scalability problem is inevitable, leading this author to believe necessity will yet again be the mother of invention.
Kevin A. Shannon
DGPO, VP Communications
 insideBigData – The Exponential Growth of Data. Retrieved from https://insidebigdata.com/2017/02/16/the-exponential-growth-of-data/
 Petrov, Christo – Big Data Statistics 2019. Retrieved from https://techjury.net/stats-about/big-data-statistics/
This quarter’s column was written by Kevin A. Shannon. Kevin is an award-winning data governance expert and architect of modern, innovative EDG solutions. His approach to Data Governance has made him a trusted advisor and valued resource to the global community of data governance practitioners. Kevin has more than 20 years’ experience conceiving and implementing enterprise data management strategies and technology solutions, defining information architecture, establishing vision from concept through completion, and overseeing the associated data, technology and business process change.