Technology plays a pivotal role in the expansion and growth of businesses, however, it’s common to find many large organisations grappling with legacy systems. Transitioning from these dated systems is no walk in the park. Cloud migrations, though beneficial, can often come with a hefty resource cost. We sat down for a quick chat with Yucel Yuksel, a Chief Data and AI Executive with 20 years of data and digital transformation experience, to find out how he approaches cloud migrations to drive technology sustainability. With a knack for driving transformation through 2 global cloud migrations, a passion for data governance, and proven success in articulating the value of data management, Yucel is a key player in the field.
In the past data strategy was overlooked even in large companies because data teams were perceived as a support function That’s why many companies failed to generate high-ROI from these investments. There was also another root cause: the focus areas of data leaders e.g., infrastructure, delivery, data modelling, etc. There is a chicken-egg relationship between the positioning of Data Leaders and their focus.
Without enough governance, self-service analytics resulted in hundreds of similar and inconsistent reports throughout the organization. Another improvement area was enterprise data modelling. Because of being a perceived cost centre and lack of business value- focused prioritization, data teams delivered the requests as quickly as possible without seeing the big picture. In one of my previous companies, I found out that the objective of some reports was only supporting the bonus calculation of an employee who had a good relationship with the data team. Moreover, in many companies, communication with technical teams and stakeholders needs to improve. In many companies, there is no solid process and well-defined roles and responsibilities for data roles. To overcome this rigidness, companies were forced to develop or hand-code user business intelligence. Over time these reports become inaccurate as changes are made to internal systems and it is unsustainable for data teams to make manual updates to potentially thousands of reports. It is the unpicking of these thousands if not millions of lines of custom code that can make cloud migrations a daunting task. For instance, if a recently onboarded executive introduces brand new Key Performance Indicators (KPIs), there is no guarantee that an existing custom-coded report will support these metrics. It could be days or weeks before that report could be created depending on the complexity of the business logic and the data model created on top of the various source systems. Multiply that by hundreds of stakeholders across varying departments and locations and it’s easy to see why it cannot be sustained without an overarching strategy. To resolve such issues, companies need experienced and visionary data leaders who can create a data strategy aligned with corporate goals, and deliver business value with impactful use cases while implementing a future-proof infrastructure.
Future expandability is one of the big pluses. In principle, the modern data stack can be supplemented at any time with services from Azure. It provides all important interfaces to ensure new data sources can be easily integrated. There are standard connectors for the integration of third-party technologies. The other benefit which many people overlook is increased security. Last but not least, establishing key user groups and communities, as well as a dedicated portal for end users are significant milestones for self-service analytics.
The company-wide analytics platform generated positive feedback from a business point of view. A reliable, integrated database has been created that allows stakeholders to make fact-based decisions at all levels of the company and in all areas of the company, thereby also intensifying customer relationships. It is already evident that we are faster and more efficient in implementing data-driven requirements. The next step was to implement real-time monitoring of machine data in order to further optimize production processes and capacity planning.
Although the people side of the story is pivotal, the technical side should be handled to ensure high scalability and the easy connection of new data sources. Thus, the main goal of cloud migration was not only to create the basis for consistent and high-performance analytics across all subsidiaries but also to enable digital use cases requiring various types, forms, and velocity of data. That’s why the biggest challenge of such projects is managing the change towards the best practices as well as the expectations of stakeholders during these long-running projects.
Yes. Imagine you are asked to produce this new report for the new C-Level Executive. You open the existing report which was poorly designed due to inefficient request management and time pressure. There are thousands of lines of code with no comments or documentation. Imagine also that the only person who holds the knowledge has left the company. It can easily turn to a nightmare for the new team members that results in frustrations, and it is also a chicken-egg relationship. Given that data engineers typically transition to new roles every two years, which often poses challenges for newcomers trying to navigate complex codebases, data leaders should consider sustainability in talent management with employer branding, impactful projects, etc. as well as ensuring enough level of documentation as text or video in a shared folder.
Thanks to medicines, lol 😊
Amid these difficulties lie opportunities for improvement. First, establishing effective monitoring mechanisms and SLAs is quite helpful. Moreover, the root-cause analysis of problems as well as transparent communication with stakeholders make everything smoother.
Migrating to the cloud not only presents a chance to address existing issues but also to enhance sustainability in the long term by modernising processes and systems. As an optimist, I believe that, while the departure of a key team member can be disruptive, it may also serve as a catalyst for innovation and strategic planning to overcome obstacles and drive progress within an organisation.
When I joined one of my employers, we were using another BI tool, but it often showed inaccurate reports. Due diligence was completed, and a decision was made to move to Power BI to eradicate a range of multi-faceted issues. This got us to a place that was good but not great and naturally you would think that further resource investment could have got us to great. However, it’s never simply a case of adding data resources to fix such issues. Without standardised business processes, empathetic data architecture design and consistent reporting cadences, it doesn’t matter how much data resource you have, you will always be fighting against an error-prone data set that is difficult and costly to maintain.
True and trusted business intelligence must start with Ownership. Ownership of systems, processes as well as standardised reporting. Only with true ownership can business intelligence be configured in a way that is future-facing and commercially astute. I’m in no way trying to downplay the importance of internal knowledge. However, when you are playing fiddle to a myriad of stakeholders, who all have their own way of doing things, and have no appetite to change, moving the needle from a BI, machine learning or analytics standpoint is highly challenging and time-consuming.
After Ownership comes Awareness. This is the Why. Without Awareness it’s very difficult to win the hearts and minds of your stakeholder groups. Thus, if you want to become a data-driven company or you want to employ emerging technologies, your stakeholders must understand that data structure has to be adapted to succeed on your journey. Therefore, the data leader should have business acumen and be able to communicate in their language.
The final part is Data Literacy which is the ability to explore, understand, and communicate with data in a meaningful way. Fundamentally this is more of a change management task rather than a Power BI training. The main objective was to expand the user groups in the company and enable as many employees as possible to work efficiently with data. This means that every employee can become productive immediately without major barriers and get a comprehensive view of data and analyses worldwide. Thus, finding data champions that will disseminate the data strategy would be a good starting point.
Hence, in essence, the success of any data-driven initiative hinges on a holistic approach that addresses ownership, awareness and data literacy. Without the backing of top-level management, initiatives to improve business intelligence and adopt emerging technologies may face significant hurdles in implementation and adoption.
Change management is a crucial aspect that should not be overlooked. This involves implementing strategies to help employees adapt to new processes and technologies smoothly. However, it is normal for data departments to have a lack of dedicated resources for this important task. While collaborating with HR and internal communications could be beneficial, it can be challenging due to their limited capacity to focus on other priorities. Without proper support for change management, the success of initiatives like cloud migration may be hindered. Therefore, change management activities like training, celebrating quick wins, data champions to advocate the new concepts, sharing lessons learned, inviting external speakers, etc. need to be implemented as part of the data strategy.
Remember, migrating to the cloud isn’t just about technical aspects; it also involves transitioning mindsets and behaviours to align with the new world of possibilities for optimal business results.
In conclusion, Yucel Yuksel, with his extensive expertise in data, AI, and digital transformation, sheds light on the challenges and opportunities of transitioning from legacy systems to cloud-based solutions. Legacy systems pose a significant hurdle due to their rigidness, lack of functionality, and the complexity of unpicking custom code, often resulting in inaccurate data and unsustainable manual updates. However, migrating to the cloud offers the chance to address these issues and enhance sustainability in the long term.
Yucel emphasises the importance of standardised business processes and consistent reporting cadences in achieving true business intelligence. He stresses that ownership, awareness, and data literacy are essential components for the success of any data-driven initiative. Without top-level management support, effective change management strategies and stakeholder management, initiatives such as cloud migration may face significant hurdles in implementation and adoption.
His advice for data professionals navigating migrations underscores the importance of change management and collaboration with HR and internal communications. Successfully migrating to the cloud requires not only technical expertise but also a shift in mindsets and behaviours to align with the new environment. By addressing these challenges holistically, organisations can maximise the business benefits of cloud migration and drive technological sustainability in the long term.
