Data Mesh: What’s In It For The Business?

(PopTika/Shutterstock)

Chief Data Officers view their company’s business strategy through a data-tinted lens. As a CDO, I was focused on delivering data-driven business outcomes (value) through a sustainable data foundation (culture and governance). We are measured by our ability to deliver the right data to the right people, fast. Occasionally, or more often than we like to admit, we are reminded that we are simply not fast enough.

What do you say when a business leader asks, “Why can’t I find or why can’t I access MY data?” There are too many data sources, too many data owners, too many compliance blockers… too many excuses. We don’t have great answers. It is frustrating, for everyone. Perhaps you have been in a similar situation with your business leaders. Our data is everywhere today and there is always new data coming tomorrow.  We don’t know where or how that new data set will be stored. There are too many unknowns.

While attempting to move all our data into one location is admirable, the ROI for a centralized single source of truth doesn’t hold up. The cost to migrate, the time to migrate and the active business disruption are very difficult to justify. The river of new data sources never runs dry, so your migration programs never end. We need to create a sustainable data analytics pipeline but are trapped in this infinite “migration” loop.

It is time to walk away from the outdated centralized source of truth approach. Ongoing data sprawl is making access to and sharing of distributed data difficult, and data sovereignty regulations are requiring data to remain within certain borders. This means that, from a practical standpoint, querying and analyzing data where it lives is more reflective of today’s realities. A Data Mesh model allows data to remain in place, empowering organizations to make that data available to business leaders in a context and format that they understand.

CDOs know they must deliver value and governance by integrating data architecture, process, and culture 

As vendors refine how to help customers on their Data Mesh journey, it’s important to look at the transformation principles and practices from the vantage point of a Chief Data Officer (CDO).

A CDO knows that they must deliver value and governance by integrating data architecture, process, and culture. It’s important to think about the data scientists, too, and how to enable them to spend more time on analytics and less time negotiating with other groups to find and access data.

Yet, it’s also vital to consider how Data Mesh stands to impact the business as a whole. The business does not care about where the data is or how it is made. The business cares about accelerating the delivery of trusted insights, decisions and actions. Data Mesh is designed for business users who are adopting a data-driven mindset, and data products are curated and described in a business language they recognize. As data becomes easier to understand, access and use, it opens up new competitive opportunities for businesses. Let’s take a look at how this is playing out in a few different industries.

Banking/Financial Firms: Undergoing Rapid Digital Transformation

Without question, banks are investing heavily in digital and mobile technologies, from AI-powered virtual assistants to 5G apps that make it easy to conduct most, if not all, of your personal and business operations through your mobile phone.

Data mesh can help unlock siloed data for fintech companies (Jirsak/Shutterstock)

Faced with fierce industry competition, leading banks are developing ways to keep their customers engaged and protected by offering the right products and services at the right time. The machine learning and AI engines that drive these solutions need data to generate optimal recommendations, but at most large banks, the data is distributed across multiple disconnected datasets. Moving all of this data into one place isn’t going to work, as projects like that take years. By that point, the target customers will have moved on to a new bank or AI-powered finance app.

By shifting to the Data Mesh approach, banks can feed their AI and machine learning models the data they need in a secure, efficient manner. And they can make this shift in a matter of months, not years. Before long, they’ll be regularly offering customers the products and services they need, when they need them.

Life Sciences & Pharma: Driving Discovery 

As COVID research has made clear, pharmaceutical companies, university research groups, and government agencies need to collaborate and share data with each other, and with various labs and testing facilities. If all of this data is stored in different siloed systems, these organizations aren’t going to be able to coordinate their efforts or share information effectively. But if companies and research groups adopt the Data Mesh approach and start to treat datasets as products that can be shared both internally and externally, the potential for new discoveries, research findings, and collaborative business partnerships grows tremendously.

SOPHiA GENETICS is a great example – the company has created a global data-sharing network with customers from over 780 institutions in more than 70 countries. They’re launching their own Business Data Mesh, which enables them to connect regions and data sources while adhering to compliance laws and regulations. Another example is healthcare technology leader EMIS. The company collects recently reported COVID-19 symptoms from a wide range of sources and makes it available to researchers in real-time. This wouldn’t be possible with a centralized, single source of truth model.

Better data management via data mesh can accelerate drug discovery (PopTika/Shutterstock)

Data Mesh isn’t merely changing things for data scientists and infrastructure teams—it’s opening up entirely new opportunities for the business.

Supply Chain Logistics: Supporting the Digital Twin

Across industries, companies are adopting virtual distribution centers and other digital tools to help them manage complex, distributed networks of assets. Supply chain digital twins simulate and project what is happening and what might happen along your supply chain. They’re incredibly valuable, but like the customer-facing AI engines at banks, they only work effectively if they can access relevant, recent data. You need to constantly feed them information to make that possible.

If your approach to feeding these models is to move all the datasets from the relevant parties into one place, you probably won’t have a supply chain to manage anymore. The whole operation will take too long. A Data Mesh that allows these parties to retain control over their datasets, while also making data easily and rapidly accessible to other groups, is a far more efficient way to operate.

Data Mesh Reimagines Business Possibilities

These are just a few examples from a handful of industries and application spaces. As Data Mesh interest and adoption grows, and more companies understand the challenges they’re facing on their Data Mesh journeys, organizations are discovering new possibilities every week.

What if you were to monetize these data products by sharing them with other companies? How could you benefit by sharing data with organizations within or outside your industry vertical? There is certainly much more to explore with Data Mesh, but the value to the business is very promising.

About the author: Adrian Estala is the VP of Data Mesh Consulting Services at Starburst. Prior to Starburst, Adrian served as the Chief Data Officer (CDO) at Shell for three years, tasked with delivering data value through an ambitious digital transformation program. He also worked as a director at KPMG for 13 years. Skilled with getting the most out of data-driven investments, Adrian is your trusted adviser to navigating complex data environments and integrating a Data Mesh strategy in your organization. His contagious passion for learning and innovation creates an engaging discussion with executives who are on their own unique digital transformation journey.

Related Items:

Starburst Backs Data Mesh Architecture

Data Mesh Vs. Data Fabric: Understanding the Differences

Data Meshes Set to Spread in 2022