Devising a Data Strategy Amid Evolving Perspectives: Insights for a Data-Led Transformation in 2023

In 2023, a leader in the organization is faced with the challenge of designing a data strategy that assists in modernizing the organization’s data science approach. The key obstacle is the decentralized management and governance of data in a conglomerate consisting of various companies and divisions. With the interest in data mesh gaining momentum, advice from technology companies, system integrators, service providers, and industry analysts vary greatly, leading to a myriad of opinions and recommendations.

Market conversations vary widely, ranging from assertions like “Data fabric will eclipse data mesh” or “Data mesh will become obsolete” to declarations like “Data mesh revolves around the four principles of domain ownership, federated computational governance, self-service platform, and considering data as a product.”


So, what should a leader do in this scenario? Let’s explore the forecast for 2023.


Data mesh is here to stay: Data mesh propels organizations to think about data from the use case and business value perspective, not vice versa. With companies advancing in digital and AI investments, the focus will increasingly shift towards creating data products driven by business value. Expect to see domain ownership define the data product context, a rise in self-service platforms, and expansion and decentralization of governance due to new data and AI regulations.


Federated computational governance needs improvement: The decentralized nature of building data products gives it a maverick appeal. But reckless creation without proper risk assessment and management frameworks can lead to failure. Effective data governance needs to be executed within decentralized teams, while aligning with global and business area policies and controls.

Domain ownership is transforming: Organizations will realize that ownership in a decentralized matrix of business and technology roles is a complex task. Ownership will become distributed and detailed, with domains changing based on multiple owners, governance, standards, context, utilization, and data sources.

Data mesh is shifting towards real-time use cases: Business intelligence and analytics are common starting points for data mesh. However, insights displayed on screens are no longer sufficient for fast-paced decision-making. Intelligence needs to be integrated into digital experiences, process automation, and partner experiences to have a significant impact on outcomes.


Data fabric technology acquires a contextual distributed orchestration layer: Early data mesh theories leaned heavily on microservice architecture and event-driven architecture. Technically, data mesh serves as the orchestration layer, ensuring that technology caters to and optimizes data and insights for the consumer.

Understanding data mesh technology is a challenge: The increasing interest in data mesh technology has led to a surge in its popularity. However, opinions differ on whether data mesh is a technology or a practice to monetize data. The result is confusion among chief architects and other stakeholders.


So, what should a data leader do? 

Remember that the four principles of data mesh are areas where you’ve already invested. Data mesh offers a framework to unify these principles, making decentralization more manageable and effective. However, as business stakeholders expand design thinking practices for digital, AI, and edge, their requirements will influence what and how you build data and intelligence capabilities. Ignore the marketing noise for now. Instead, focus on designing a data strategy that balances partners, practices, and platforms to drive customer experience, value, and outcomes.


More information about AI? Look here!

What is AI?

AI for Executive

What is GEN AI?

AI in Business

AI Enterprise Strategy

FAQ ChatGPT