A methodology for creating business value in data science projects for banks
Banking Data Science Canvas
Data science projects are inherently complex. Many different aspects need to be considered, ranging from infrastructure, software and data engineering, design to data modelling. All of these aspects need to come together in the right way in order to achieve the core objective: To solve a business problem for certain users and to create business value. Trying to keep things as simple as possible, always with the focus on the business problem we want to solve, is key. The Data Science canvas is our tool to achieve this.
Insight & Action
The data science canvas consists of 9 elements which are essential in a data science driven project or use case:
- Business Problem
- End Users / Personas
- Value Drivers
- Project stakeholders
- Data Solution
- User Journey
- Data modelling
- Data requirements and pipeline
- Software architecture
Our approach is drawn from principles of “Design Thinking”. Together with key representatives of the client, we identify and define the essential aspects of the project in a series of workshops. We start with formulating the business problem, defining end users (personas) working with the solution and how they and other stakeholders shall profit from the solution (value drivers). This forms the basis the next step in the process: The research phase where we explore possible data-driven solutions for the use case, describe the user journey and how end users interact with the system, look at possible data modelling and data pipeline requirements as well as assess infrastructure and software architecture aspects. We then collect and prioritize our findings and summarize our chosen approach in the canvas. The canvas is updated iteratively whenever new insights are found during the implementation phase of the project. As a project evolves, so does the content of the canvas, but the key information is always on a single page.
Ongoing communication and a common understanding of the (business) objectives are vital in complex projects. With the Data Science canvas we apply a methodology to establish such a common understanding. It allows to communicate to project stakeholders the essence of a project in a simple way and keeps all project members focused the core objectives.