When to use Data Mesh: C-suite’s view

When to Use a Data Mesh? 

Large Organisations with Multiple Business Units 

If your organisation has multiple business units that operate independently, a Data Mesh approach may be suitable. Data Mesh enables each business unit to manage its data independently, ensuring that the data produced is tailored to its specific needs. It also allows for cross-functional collaboration, enabling teams to leverage each other’s data products to achieve better outcomes. 

When Data is Critical to the Business 

For organisations whose operations rely on data, a Data Mesh can be a valuable approach to managing their data. A Data Mesh allows domain experts to take ownership of their data, ensuring that it is accurate and relevant to their needs. This approach improves data quality and reduces the risk of data errors, which can be costly to the organisation. 

When Teams are Struggling with Data Silos 

Data silos can impede collaboration and result in inefficiencies in data management. Adopting a Data Mesh approach can help break down these silos, enabling teams to work with data in a self-serve manner. It promotes cross-functional collaboration, enabling teams to leverage each other’s data products to achieve better outcomes. 

When Data is Changing Rapidly 

Organisations that work with rapidly changing data may benefit from adopting a Data Mesh approach. The autonomy and ownership of data by domain experts allow for more agile and responsive data management. It also allows for quicker experimentation and prototyping, enabling teams to iterate on their data products and achieve better outcomes. 

 

Data Mesh and how it can improve profitability, drive revenue growth, and monetisation 

As a Chief Data Officer or Chief Information Officer, you will appreciate the importance of exploring innovative data management approaches such as Data Mesh and leveraging them to improve profitability, drive revenue growth, and monetisation.

Here’s how you could consider the use of Data Mesh and its potential benefits: 

• Understanding Data Mesh: First and foremost, you need thoroughly educate yourself and the organisation about the principles and concepts of Data Mesh. This includes understanding its decentralised architecture, domain-oriented approach, and focus on data product thinking. 
• Assessing Data Complexity: Evaluate the complexity and diversity of your data landscape. If you have a large organisation with multiple business units, each with distinct data requirements, Data Mesh could be a suitable approach to manage data as a product within each domain. 
• Identifying Data Domains: Identify the different data domains within the organisation. These could be specific departments, business units, or functional areas that have unique data needs and can be treated as separate domains. 
• Establishing Domain Ownership: With Data Mesh, Facilitate the establishment of domain ownership. This means assigning accountable individuals or teams within each domain who have the authority and responsibility to manage the data products within their respective domains. 
• Data Productisation: Encourage the transformation of data assets into data products that are tailored to meet the specific needs of each domain. By treating data as a product, we can enhance its value, usability, and adoption within the organisation. 
• Empowering Domain Teams: Enable domain teams to take ownership of their data products, including data quality, data pipelines, data governance, and data monetisation opportunities. This empowers them to innovate, experiment, and deliver data-driven solutions that drive profitability and revenue growth. 
• Collaboration and Interoperability: While each domain is responsible for its data products, emphasize the importance of collaboration and interoperability between domains. This ensures that data can be shared, integrated, and utilised effectively across the organisation whenever necessary. 
• Data Monetisation Opportunities: Work closely with domain owners and business units to identify potential data monetisation opportunities. This could involve leveraging data assets to create new revenue streams, such as selling data products externally, offering data-based services, or partnering with other organisations for data collaborations. 
• Data Governance and Compliance: As a Chief Data Officer, ensure that data governance and compliance practices are integrated into the Data Mesh approach. This includes defining data governance policies, ensuring data privacy and security, and adhering to relevant regulations and standards. 
• Performance Measurement: Finally, establish key performance indicators (KPIs) to measure the success and impact of Data Mesh implementation. These could include metrics such as increased revenue from data monetisation, improved data product adoption, enhanced data quality, and domain-specific business outcomes. 

By adopting a Data Mesh approach and effectively implementing its principles, we can improve profitability by empowering domain teams to drive value from their data products, explore new revenue streams through data monetisation, enhance collaboration and interoperability, and ensure efficient and effective data governance. 

 

An esynergy data mesh case study 

One of the top companies in the world for wealth management, asset management, and asset servicing is Northern Trust. With 23 sites worldwide in North America, Europe, the Middle East, and Asia-Pacific, Northern Trust is headquartered in Chicago. 

Innovation is essential to be competitive, which is why Northern Trust bases its strategy on data and analytics. 

In order to empower its staff and clients, Northern Trust is on a next-generation data modernization journey and seeks to create a data-driven culture. This endeavour is focused on a data mesh framework that serves as the data delivery foundation for a number of applications, such as an easy-to-use self-service data marketplace. The market and framework are predicated on the principles of data as a product, embedded governance, security and business-driven IT. 

Northern Trust hired esynergy as its technology partner for data modernisation in early 2022 with the primary goal of hastening the creation of the data mesh framework. 

Modern data mesh architecture would support business ownership and IT enablement, enabling Northern Trust to handle difficult data challenges from a product perspective. 

A federated data ecosystem with best-in-class software engineering techniques implemented on public cloud was created as a result of the collaborative work. 

• establishing a multi-platform infrastructure layer with inbuilt security and access control in a mesh configuration 
• processes for data governance and management can be made more efficient and automated, allowing businesses to own their data. 
• the use of DevSecOps tools 
• making CI/CD pipelines automatic 
• finding ways to make advanced analytics provide value for the business 

 

“Northern Trust had a diverse set of stakeholders that were keen to engage in the data product strategy. By partnering with an initial three business areas spanning both Wealth Management and Asset Servicing we were able to build a second-generation data mesh that met their specific needs for autonomy, isolation, and speedy time to market. This has led to data products rapidly going into production with an expanding pipeline of new data products being onboarded into the data mesh ecosystem.”  – Sunny Jaisinghani – Data Mesh Delivery Lead, esynergy

In order to democratise data and strengthen analytical capabilities, Northern Trust can use its platform for data as a product to increase the consumption of various data assets, both internal and external. In a conventional strategy, finding the appropriate data can account for up to 80% of the whole project time for analytics, with extract, transform, and load (ETL) adding additional weeks or months. Northern Trust can now complete these initiatives in a matter of days or even hours with less data risk because to the deployment of the new data mesh pattern. For instance, in the first use case for data products, analytics were used to boost the effectiveness of a Fund Accounting operations team. Machine learning models were built to increase the correctness of exceptions forwarded for manual validation. 

This resulted in fewer overall exceptions, quicker results validation, and improvement of SLAs, and reduced risk of error from manual interventions. 

According to Kelley Conway, Global Head of Corporate and Digital Strategy at Northern Trust, “Data mesh is an evolutionary leap forward, utilising cloud-based tools and modern technology to liberate data for use by our clients and internal partners.” “Our clients should have access to any information they require, whenever, wherever, and however they choose. By reaching clients where they are and distributing data through portals, APIs, and cloud-based apps while guaranteeing data consistency, quality, and control, we have developed a foundation that enables optionality for customers. 

The Data Mesh pattern has also been applied to add data to a cloud-based Analytics Workbench that is powered by Databricks and enables large-scale, quick model creation. This technology has been used by a team of data scientists to speed up business-driven advanced analytics and produce outcomes that will enhance client experiences, increase cross-sell opportunities, and save operational costs. 

The advantages continue, though. Along with early analytics successes, Northern Trust is developing a consumer data marketplace to enable grow more effectively. To start, imagine an internal “Google-like” search engine where employees can find data products with all the necessary information and also give feedback to data product owners to help them get better. 

In conclusion, adopting a Data Mesh approach is not always the right decision for every organisation. However, for large organisations with multiple business units, those whose operations rely on data, struggling with data silos, or dealing with rapidly changing data, it can be a valuable approach to managing their data. By emphasizing the autonomy and ownership of data by domain experts, a Data Mesh approach promotes collaboration, improves data quality, and reduces the risk of errors, enabling organisations to achieve better outcomes.