Data Talks #1 - "How to implement a data governance framework" - Shamma M. Raghib
Written By: Jérémy Corbet
Back To All Blogs
Here is the first article of my new blog series in which I will be interviewing Data Leaders. The idea is to share knowledge and updates on what’s happening in the Australian data market.
Today we start with Shamma M. Raghib who gives us her insights on the hot topic which is data governance.
Who is Shamma?
Shamma is an experienced Senior Presales Consultant in Data Management and Data Governance space with a demonstrated history of implementing Data Governance frameworks and working in complex data management transformation projects including Data Strategy, Data Governance, Data Migration, Contact Data validation, ETL, Data Matching and De-duplication, Single Customer View / Customer 360 view.
She graduated from Collibra's Chief Data Officer Summer School in 2018. While all things data are her core strengths, she also has a strong background in business development. She also followed and graduated in FinTech: Future Commerce focused around Blockchain and Future Finance Infrastructure from Massachusetts Institute of Technology.
Currently she is working in Experian as a senior data management consultant, helping organizations make best use of their data and implementing data quality frameworks.
A lot of companies struggle to implement their data governance strategy – misalignment with business stakeholders, lack of C-level support, technological challenge, etc. In your opinion, what are the biggest challenges that you have experienced or seen in that regard?
First and foremost, you need to understand what is driving your program, because I have to tell you, I have never met an executive who woke up and said, “Let’s fund data governance”.
Secondly, quite often the value of the data is not being fully conveyed to the business. What data value is assigned to a dataset, in turn, affects the policies around that data. If the value of data is not understood properly it is very likely that the policies will be affected and very likely the governance around it will result in project failure. A good tool is necessary to solve the problem; however, a good tool does not guarantee a good governance program.
Quite often organizations also think that it is a technology problem and the tech teams clutter together to form a governance committee and mandate a technology to “solve” the problem. Unfortunately, this is still quite typical in relatively mature markets like Europe.
Quite often, solution architects have cross company overview around data and often this may lead the IT teams taking ownership of the governance. However, this is a wrong approach since most policies and processes around the data come from the business. So, what role does IT play? They can evangelise the technology and vet the appropriate tools to meet the business requirements.
Business stakeholders often mention data governance as another obstacle. How do you handle the relationship with them to make it scalable across the business?
If business stakeholders and data consumers are constantly asking questions such as, “Where does this field come from”, “Is this data accurate?” or “Who owns this data and can answer my questions?” then they have little time to achieve their actual business goals. The best way to approach this issue is to educate, but not just educate. The mandate of being “data-driven culture” has to come from both top down and supported by bottom up. Most data governance programs involve a significant level of change management and transformations within business processes; how the organizations perceive the value of data as well how to take ownership of data.
What are the emergent trends that you see in this industry?
There are many emergent trends in data and tech world now including the typical buzzwords around AI and Machine Learning (genuinely feel people need to stop using buzzwords) + any data management domain. One that is to look out for is Explainable AI – which is basically a way to simplify AI algorithms for humans. For example, if you want to make users understand why a duplicate record that looks similar is not a duplicate – the algorithm is sophisticated enough to maybe understand why however a human explanation might be a visual way of representing the problem.
The second trend to look for is around Hyper Automation - sophistication of the automation (intelligent data discover, intelligent analysis, collaborative design, business centric measure, result centric monitor, reassess iterations, and automate.). Automation alongside AI with IoT capabilities (lovingly known as Edge AI) will soon be used across IoT industries – this can already be seen in projects with companies like Ford.
Alongside these, I do believe practically speaking, blockchain is becoming more and more the new norm for a lot of technologies specially around supply chain, risk and identity use-cases with dynamic smart contracts. In near future blockchain will be implemented with AI and ML considerations to make it scalable.
In terms of recruiting data governance professionals, what does good look like to you?
This is a tough one, however, I do believe a good data governance professional does not only keep up to date with the recent policies and regulations that affect how data is handled in an organization; but is also able to do change management. 70% of the work in data governance program is actually around change management and convincing the right stakeholders that governance has become a necessity rather than a “Good to have”. A good data governance professional must manage people relations well but also be able to mitigate risks by providing valuable data strategy guidance for teams.
Is there anything else you would like to share with the data community?
If you need help with your data governance and data quality programs do not hesitate to connect with me or follow me at: https://www.linkedin.com/in/shammaraghib/