Data Mesh for ESG

Climate change is one of the most pressing challenges faced by humanity. Today, customers, regulators, and governments demand climate action and business accountability. As a result, organizations are increasing investment in environmental, social and governance (ESG) initiatives and taking steps to make their efforts as transparent as possible.

It’s imperative that most recognize and are more than ready for a response. But, for millions of teams across the globe, there’s one big question still limiting their ability to take the right action: How? 

According to IDC, spending on ESG business services will grow to $158 billion by the end of 2025. That forecast confirms something significant. Not only are organizations ready to invest in sustainability, but many are also so unsure of the best ways to go about it that they’re willing to pay heavily for support and guidance.

In a perfect world, organizations should be able to use their data to identify and prioritize areas in need of investment intelligently and where change is needed most. But for most teams, their data isn’t accessible, complete, or strong enough to help them do that effectively.

A recently published study from the Queen’s University’s Institute for Sustainable Finance in Canada found that just 6% of organizations are “very satisfied” with the availability of sustainability data, and 41% say obtaining additional data is their top priority.

Under current architectures and structures, obtaining high-quality, usable sustainability data is easier said than done. But a new decentralized approach to data architecture developed by Thoughtworks — Data Mesh — can potentially push things in the right direction.

What makes Data Mesh a good fit for sustainability?

Data Mesh is a decentralized data architecture and operating model where data is a product and the teams closest to that data own it — rather than all pooled together into a single centralized lake or warehouse.

From a business perspective, it helps cut out many bottlenecks teams have seen with centralized architecture approaches. For example, instead of going to a central data team when they need insights, groups can help themselves to what they need and when they need it — dramatically decreasing time to value.

From a sustainability data standpoint, Data Mesh provides a solid potential operating model and governance approach. Each domain is responsible for managing and curating its own data products within a Data Mesh. They’re empowered to collect and use data their way, but to ensure that their data products have value for teams across the business, some cross-functional responsibilities must be upheld.

Let’s take Scope 3 emissions as an example. Right now, reliable Scope 3 emissions data is tough to come by for most organizations. That’s primarily because the business hasn’t bought into the value of gathering it. As a result, there’s no direct incentive or requirement for the teams closest to emissions data (the ones doing things like booking travel or outsourcing services) to capture or maintain it.

Driving sustainability in enterprise-scale organizations requires cross-functional effort. Setting cross-functional requirements can help teams clearly understand how they can help drive sustainability through their decisions, actions, and the data they collect and manage. A Data Mesh makes it easier to set and uphold requirements at scale.

Within a Data Mesh, sustainability can be defined as a critical cross-functional use case for data relating to suppliers, travel and accommodation providers, or any kind of outsourced service provider. That helps the teams closest to that data better understand its sustainability implications and what they can do to help capture and maintain high-quality, reliable sustainability data.

Give sustainability a seat at the data governance table

Cross-functional responsibilities like that are just one example of federated governance — one of the core pillars of the Data Mesh paradigm. In a Data Mesh, governance responsibilities aren’t just defined by a central data team. Instead, everybody invested in specific data outcomes gets a voice to ensure that everyone’s data requirements are met and upheld across the mesh.

It enables productive conversations where every domain, and the organization at large, can start to better understand how its data can inform sustainability and ESG efforts. It doesn’t create high-quality sustainability data immediately, but it enables the understanding and process changes required to generate more of it long-term.

Rise to new reporting demands and start automating sustainability

Data Mesh isn’t just great at creating cultures that naturally gather more high-quality sustainability data either. It can also help teams report on that data more efficiently and increase their operations’ transparency.

Within a Data Mesh, domains can create their own data products to serve specific data use cases. Standardized templates and guardrails for creating those products help ensure that the data within them is highly searchable and discoverable. So, when a regulator or any other stakeholder wants to look at your sustainability data, it’s far easier to find and provide it to them.

But reporting is the beginning of what organizations can do with that data. By making data highly available, standardized and of reliably high quality, a Data Mesh can help teams access the data they need to start exploring ways of autonomously optimizing sustainability.

For example, models and algorithms could be used to detect anomalies and outliers in sustainability data as they’re recorded, enabling immediate automated actions to be taken, helping to reduce emissions at their source.

Data Mesh has high potential, but it isn’t a silver bullet for sustainability

While those changes have the potential to push your sustainability efforts in the right direction and help you tackle enduring challenges around the accessibility of high-quality sustainability data, it’s important to recognize that Data Mesh isn’t a silver bullet. It isn’t going to solve every challenge immediately.

Building a Data Mesh and applying it across an organization takes a major commitment to change. With it comes huge changes in how teams work, new responsibilities, and potentially even shifts in how domains are structured. And even if your organization is ready to make that commitment, Data Mesh isn’t a suitable architectural approach for everyone.

We’re still in the early days of the Data Mesh paradigm. But results from some of Thoughtworks’ live Data Mesh projects have been extremely encouraging.

At Roche, we’ve seen how federated governance across domains can impact data quality and how high-quality data can be rapidly translated into business value. Similarly, at the Department of Veterans Affairs (VA), we’ve seen how purpose-built data products can improve human outcomes and lives for thousands of people.

Direct examples of sustainability use cases within Data Mesh are limited today. But from what we’ve seen, the potential is certainly there. We’re constantly exploring new use cases and areas where the principles of Data Mesh could help tackle significant business and societal challenges. Today, sustainability is one of the brightest areas for innovation and experimentation.