The professional world’s relationship with “Data” is generally based on a love-hate dynamic, and incomprehension in the majority of cases. We’re in the midst of a professional era where technology and Data reign supreme. No one doubts that the optimal structuring, harmonization and analysis of data are key factors in the success and survival of businesses in today’s world. This is even more true in highly regulated fields such as financial services, where the lack of effective monitoring of activities can prove fatal to the reputation and business model of these institutions.
And despite this paramount importance, data management in most cases generates frustration and anger among management teams and staff. Indeed, very few companies manage to set up an optimal IT structure and unite their teams around a culture focused on the benefits of understanding and using the data in their possession.
Such is the inevitability surrounding the implementation of new software, business processes and systems that it has given rise to a new managerial concept: “Post-implementation crisis management”. Teams work on the assumption that most projects are doomed to failure, and that the hassle and cacophony generated will have to be repaired by setting up a new team of specialists.
I believe that almost all projects fail for one simple reason: the biased perception of management and decision-makers who believe that this field is the exclusive domain of IT and technological skills. Based on this belief, most major projects are entrusted to the IT, Development and Data Scientists teams, ignoring the most important aspect: the involvement of Business teams in managing these programs.
For my part, I’m firmly convinced that setting up multi-disciplinary teams combining operational, IT and technology teams is the key to success for maximum return on investment from data analysis and utilization.
The Cross Industry Standard Process for Data Mining (CRISP-DM), the most widely used process model in data mining, makes it abundantly clear: data is there to help understand and analyze the business. Understanding the concrete questions to which we are looking for answers is the foundation of any analytical process, and all projects must start with the same question: “What is my business problem? How can the data I have at my disposal help me see things more clearly?
Whether it’s analyzing customer transactions, satisfying KYC/AML regulations, managing regulatory obligations or gaining an in-depth understanding of customer needs in order to adapt their business strategy, financial institutions need to implement an inclusive, cross-functional approach to their staff, enabling them to contribute their expertise and work hand-in-hand with technology teams on the technical aspects of the data analysis models they need. This is what I call the implementation of a Data-centric organization, where the added value of data is recognized by all, and the entire staff is continuously mobilized to get the best out of their data.
Following on from CRISP-DM, I believe that the path to becoming an organization with a true “Data” culture must follow a cultural transformation based on three axes: Strategy, Expert Teams and Technology.
Each project starts with your needs and objectives: “I want to analyze my customers’ transactions and identify potentially fraudulent operations” “I want to structure my customers’ data for my KYC reporting” “Which customer segments are most profitable for my business? Which customer segments are most profitable for my business?” In the second stage, the Business teams will look at the data they have and see how it can be analyzed. This step provides clear information on how you’re going to take advantage of them, and explains to Data Scientists and IT how they’re going to model the process.
2. Teams of experts
The second step is to define the appropriate actions to launch the active collaboration phase. This can only be done by preventing what I call “data communitarianism”, where technology and non-technology teams fail to communicate seamlessly and optimally. This issue must be recognized by management and dealt with appropriately. One of the solutions for defining total collaboration around data is to create cross-functional teams comprising Data Scientists, IT, Sales, Product Management and Operational teams for a global view of the organization. This new organizational structure will enable total emulation between teams during the process of understanding and preparing data.
This last step is also critical to the process. Data Scientists and Analysts who are modeling need to share their progress interactively with business teams to make sure they’re on the right track, and adjust when necessary. The Modeling and Evaluation stages are anything but a rigid straight line. Numerous adjustments are required to achieve the most optimal and satisfactory model for successful deployment.
Data analysis and understanding represents a unique opportunity for banks, asset managers or insurers to meet their obligations, adapt their strategy and ensure their long-term survival. When business specialists, operational and technology teams work together, these projects generate descriptive and predictive analyses that are understandable, powerful and satisfying for everyone.
Dear users, on 15/06/2022 Internet Explorer will be retiring. To avoid any malfunctioning, we invite you to install another browser, such as Google Chrome, by clicking here, or the one of your choice.
Please check this before contacting us in the event of a problem.