15. July 2024

From CDO to Data Success Board: how to successfully transform into a data-driven company

Gaining data-based insights and making decisions, optimizing processes with data, using (generative) AI: there are plenty of companies that want all of this. You occupy the role of chief data officer (CDO), the person who is supposed to set up and implement a suitable data strategy - and in many cases this fails. But why? Because this subject is almost impossible to tackle alone. In this article, we will show you why data is better approached as a team sport and how the "Data Success Board" helps to anchor and implement the data strategy in the company over the long term.

the company over the long term.

Without data, we have to resort to guesswork… The fact that data is of key importance in our increasingly digital world has been well known for a while now. Companies need this data to analyze customer needs and derive sales measures in order to optimize products and services and make processes more efficient. However, for a long time, the data was housed in the controlling department. Nonetheless, this didn’t stop the data from showing up in various other departments and staff units. Most companies have now attained a level of digital maturity that comes with an understanding of the importance of end-to-end digital and data-driven processes. Data should be accessible and usable in all processes, across departments, and systems. Not least, this ensures that new technologies, such as generative AI, can be effectively utilized.

The crux of the matter is that, in many cases, the concrete vision of a data-driven company remains unclear. Historically developed application landscapes, data silos, and disruptions in media and processes often make it difficult to achieve this target vision. The data landscape and data streams must first be documented, consolidated, and streamlined. Only then can specific use cases for improving data utilization be identified and implemented. This, in turn, requires expertise in new technologies such as machine learning and generative AI, as well as “data literacy” as a general competency.

Case

Data Competence Training

Company track: Basic knowledge for all employees
Business track: Data competence for specialist departments
Operations track: Technical know-how for IT specialists & analysts […]

“It takes a village to raise a child”

The saying implies that successful, holistic education requires the combined experience of many individuals. This also applies to the data strategy. After all, how can one person alone be expected to have an overview of all departments, processes, tasks, associated systems, and data, and find optimal solutions for all use cases? It requires the perspectives of specialist departments, the tech expertise, and the insights of management. Data cannot simply be added as a new organizational area alongside business and IT; it is a cross-cutting issue that impacts all areas of the company, involving everyone.

From CDO to Data Success Board

What can it look like in concrete terms if “everyone” is involved? In our experience, one effective control model that has demonstrated success in practice is the Data Success Board. It ensures that the topic of data is anchored at the C-level and integrated into the strategy process. Initially, members include representatives from management who provide business and IT perspectives, often including the chief data officer (CDO), who offers a centralized view of data to the committee. Because of its composition, the Board maintains close links with relevant stakeholders in the organization, such as:

Works council or staff council
Compliance officer
Data protection officer
Representatives of the departments

They all contribute their experience in managing data. They understand where processes falter and tasks become more challenging due to missing, duplicate, or outdated data, and where simplification of data maintenance is needed. Together, they can discuss the need for new solutions and process improvements, and create synergies in the process. And the role of the chief data officer? It remains. Rather than addressing the data strategy issue alone, it is now distributed across multiple shoulders. The CDO facilitates the board discussions and assumes the lead role in implementing the strategy.

Moving forward with implementation with the Data Success Board

As a committee, the Data Success Board coordinates the strategic management of data and develops the vision for transforming into a data-driven company. For this vision to be effectively operationalized, a sustainable structure is essential for its implementation. The Data Success Board rests on five operational pillars that have proven effective. Each pillar represents a distinct strand of work within the overall project, each with its own designated leader. High-level horizontal networking between divisions is a critical success factor in implementation.

In this area, the focus is on developing and implementing a clear and comprehensive roadmap for use cases. It should align with the company’s strategic goals and include potential use cases that generate added business value, with costs justified relative to this added value. Examples include improving the recognition rate of customer inquiries in insurers’ input management, optimizing fault rectification in production using generative AI, and automating the processing of IT tickets. Companies in every sector and of all sizes have the potential to leverage data across their entire value chain. Best practices for implementing data, machine learning, and AI projects offer additional guidance on what is achievable and realistically feasible.

Three key questions to start with:

What business goals should be achieved with data-driven solutions?

Which potential use cases can help to achieve these goals?

Which specialist requirements must be taken into account in order to successfully implement these use cases?
Here, the emphasis is on developing a resilient and scalable data architecture that serves as the foundation for data-driven decision-making. This involves defining data models, choosing appropriate database technologies, and implementing data pipelines for collecting, storing, and processing data. Data management also involves developing strategies for ensuring data quality, integrating data seamlessly, and maintaining robust data security measures. This is essential for using the necessary data reliably, consistently, and securely. All of this requires appropriate technical expertise and is correspondingly complex, depending on the size of the company and its IT landscape. The use case roadmap outlined above offers an overview. It establishes the criteria for which use cases are to be implemented and when, as well as specifying the data sets that need to be established and expanded in terms of quality, quantity, and provisioning speed. This leads to a data roadmap that closely aligns with the use case roadmap and ideally adjusts in an agile manner.

Three key questions to start with:

How effectively does the company’s current data infrastructure meet the requirements of the use cases and the overall target vision?

What data models and standards are necessary to ensure consistent and high-quality data management across all defined objectives and use cases?

What data pipelines and processes should be implemented to efficiently manage the collection, storage, and processing of data?
This area focuses on establishing and managing the infrastructure – both technical and organizational – for data analysis, machine learning, and AI. This involves implementing tools and platforms for data visualization, analysis, and model development, as well as establishing MLOps practices to automate and scale ML workflows. However, value is only realized through the operationalization of data and AI services, which means integrating them into operational business processes. For this to be a success, the infrastructure must be designed to be flexible, scalable, and reliable to meet the requirements of both current and future use cases. This enables effective utilization of data and analytics within business processes.

Three key questions to start with:

How well do the current tools and platforms used for data visualization, analysis, and model development meet the requirements of the use cases?

How can an infrastructure be set up that is flexible and scalable and enables the effective use of data and analytics?

What MLOps practices are appropriate for enhancing the automation and scalability of ML workflows and optimizing model performance?
This pillar emphasizes cultivating and expanding a data-driven culture with corresponding data skills, often referred to as “data literacy,” within the company. This involves training and developing employees and fostering an environment that encourages collaboration and knowledge sharing. Fostering an open and learning culture raises awareness about the significance of data and enhances employees’ capability to make data-driven decisions and innovate solutions. This empowers them to directly contribute to enhancing their day-to-day work and the overall success of the company.

Three key questions to start with:

What is the existing data culture within the company, and how strong is the awareness of the significance of data and analytics?

Which competencies and skills should be cultivated to facilitate making data-driven decisions and crafting innovative solutions?

How can an environment be created that encourages collaboration and knowledge sharing around data and supports an open and learning culture?
This includes establishing and executing guidelines, processes, and control mechanisms to guarantee data security, data protection, and compliance. Simultaneously, implementing appropriate guardrails is crucial for gaining acceptance of AI- and data-driven products and services among both end customers and the company’s employees. Specifically, this involves developing guidelines for data classification and access control, ensuring compliance with legal regulations such as the EU GDPR and the EU AI Act, and considering ethical aspects in data handling. Creating a robust data governance framework guarantees data integrity, confidentiality, and availability, bolstering stakeholder confidence in data utilization.

Three key questions to start with:

What current guidelines and processes have been implemented to ensure data security, data protection, and compliance?

How can these guidelines and processes be improved and adapted to the constantly changing regulatory requirements?

How can we raise awareness of ethical aspects when handling data and ensure that data-based decisions are made responsibly?

Anchoring & operationalizing data strategy sustainably

The introduction of a Data Success Board and the implementation of the five pillars offer companies a structured approach for successfully implementing the transformation to a data-driven company. However, the journey to becoming a data-driven company is not a one-time effort; it is an ongoing process that demands continuous development and adaptation. The set of methods presented here provides exactly the right basis for this.

If you would like to discuss the Data Success Board and the topic of data strategy in more detail, please contact Dr. Markus Knappitsch and his colleagues: You can contact them here.