Mindset & Structures
Data-driven transformation is primarily about harnessing the power of data to create value within your specific business domain. As a domain expert, you play a crucial role in this process. You are the architect who defines the data-driven value. You identify the information bricks - the specific pieces of data that you need to generate insights and drive decision-making.
These requests are then forwarded to the centralized information factory, which is a metaphor for the data management and analytics function within the organization. This function prioritizes the requests and determines which information bricks to build first. As the company matures in its data capabilities, the process of obtaining the necessary information bricks becomes faster and easier.
The key to success in this data-driven transformation lies in systematically translating your domain expertise into data-driven projects. This approach increases the odds of success by ensuring that the data and analytics initiatives are closely aligned with business needs and objectives.
Chapter 1: The Mindset of a Data-Driven Architect
To successfully drive data-driven transformation, you need to adopt the right mindset. This mindset revolves around three key principles:
- Being Business-Centric: The focus should always be on the business value that can be derived from data. This might seem obvious, but it's easy to get caught up in the technical aspects of data and lose sight of the business objectives. A case in point is a German car manufacturer that was able to identify a critical factor for predicting the lifespan of vehicle batteries - the heat of the battery - which was overlooked by a data-centric team because there was no existing sensor data on this factor. By taking a business-centric approach, the executives made a compelling case for investing in additional sensors to collect this data, thereby potentially extending the lifespan of the batteries and saving the company money.
- Adopting Design Thinking: This is a problem-solving approach that involves testing assumptions and iterating on ideas, rather than spending a lot of time and effort on detailed business plans that may not hold up in the face of reality. The focus is on learning as quickly as possible what works and what doesn't, and then iterating on the idea based on these learnings.
- Seeing Data-Driven Innovation as Engine for Transformation: Even though the success rate of data-driven innovations is relatively low (between 3-5%), they can serve as powerful catalysts for transformation. When business domain experts design data-driven innovations and systematically test key assumptions, they can identify areas where the organization needs to transform in order to realize the potential value of these innovations.
Chapter 2: Organizational Structures for Data-Driven Transformation
A supportive organizational structure is a prerequisite for successful data-driven transformation. This structure should consist of four levels:
- Sprint Teams: These teams are made up of business experts who develop the ideas. They do not include data people, because the focus at this stage is on identifying the business needs and objectives, not on the technical feasibility of meeting these needs with data.
- Support Functions: These are the data, IT, and research experts who provide on-demand support to the sprint teams. They help the teams understand what is technically feasible and how to leverage data to meet their business objectives.
- Enabling Functions: These include a coach who coordinates all sprint teams and provides methodological support, and trust experts who provide guidance on issues such as ethics, data privacy, and security. The trust experts are particularly important because they help ensure that the data-driven initiatives are not only technically feasible but also ethically sound and compliant with relevant regulations.
- Decision Board: This is the body to whom the sprint teams pitch their ideas. They decide which innovations to sponsor and which investments to prioritize to advance the transformation and reduce barriers for successful data-driven innovations.
Chapter 3: The Process of Data-Driven Transformation
The process of data-driven transformation involves eight steps: ideation, design, exploration, testing, iteration, validation, reflection, and pitching.
During the ideation stage, the aim is to translate domain expertise into innovation objectives that can be tackled through data and AI. The design stage involves specifying one idea from a business and data perspective and prioritizing key assumptions to test. The exploration stage involves understanding key assumptions, which are then tested in the testing stage. The iteration stage involves redesigning the idea based on the learnings, and the validation stage involves anticipating what evidence the jury expects for the redesigned idea. The reflection stage involves taking a step back from the idea and reflecting on organizational learnings. Finally, the pitching stage involves seeking sponsorship for the idea and discussing transformation needs with the decision board.
Key Learnings:
- Data-driven transformation is about leveraging domain expertise to create value through data.
- The mindset of a successful data-driven architect involves being business-centric, adopting design thinking, and seeing data-driven innovation as an engine for transformation.
- An effective organizational structure for data-driven transformation involves sprint teams, support functions, enabling functions, and a decision board.
- The process of data-driven transformation involves eight steps: ideation, design, exploration, testing, iteration, validation, reflection, and pitching.
- Despite the relatively low success rate of data-driven innovations, they can serve as a catalyst for transformation.