×

In 2019, Zhamak Dehghani defined a new approach to manage data based on decentralized ownership of data. As a refresher, Data Mesh is based on 4 tenets:

  1. Domain driven ownership
  2. Data as a product
  3. Self-Serve Infrastructure Platform
  4. Federated Governance

A comprehensive definition of Data Mesh is beyond the scope of this post and you can read the book for free, here

Like many data analytics professionals, we’re excited about the promise of Data Mesh, however, we believe there are a few implications of this transformative strategy that will impact the role of the Chief Data Officer (CDO).

With Data Mesh, domain ownership of data means decentralizing the ownership of ingesting, transforming, and serving of data to domains (business units or functions). From an organizational perspective, it would initially seem that this removes or drastically reduces the need for a central data team. This means that these domains are responsible for the activities related to data management such as information security, data quality, data lifecycle management of the data owned by the domain as well as data that’s exposed beyond as data products.

Enterprise-wide governance, security and quality policies are defined by the federated governance function within a Data Mesh. This function is composed of Domain Owners and the self-service infrastructure platform teams (that typically report into the Chief Technology Officer). The responsibility for implementing these policies rests within the domains, and the responsibility for automating this policy implementation and for policy validation resides with the self-serve infrastructure platform team.

In the past, when we have discussed this particular arrangement, others have argued that Domain Owners aren’t senior enough to make these decisions. In situations where domains are based around small numbers of employees, that may be the case. 

In reality, in larger organizations, where Data Mesh is more appropriate, it’s likely not the case. In a multi-national financial institution which has 500+ employees in the Anti-Money Laundering function — a domain that the overall organization has defined — the Domain Owner would no doubt hold a very senior position in terms of organizational responsibility, and so would definitely be senior enough to define and be responsible for implementation of policies.

Before we take a closer look at the implications of Data Mesh adoption in relation to the CDO, it’s worth defining the CDO role. In 2015, Gartner defined the role of the CDO as a senior executive who bears responsibility for information protection and privacy, information governance, data quality and data life cycle management, along with exploiting data assets to create business value.

In some organizations, the CDO role extends far beyond this set of responsibilities, and some, it’s less, but the Gartner definition is a reasonable baseline definition.

If we then consider what we suggested above about where responsibilities lie within Data Mesh, all the existing responsibilities that fall under a CDO become decentralized into domains, even ‘exploiting data assets to create business value’.

While this sounds like heresy, the CDOs that we have canvassed with this perspective have given us a range of responses. However, most have agreed that many, if not all of their current responsibilities will be decentralized to the domains.

One CDO who has driven a very successful adoption of Data Mesh within his organization asked rhetorically, “We don’t have a Chief Laptop Officer, so why should we have a Chief Data Officer?”

Taking all of this into account, it’s likely that businesses will reconsider the role of the CDO and might wonder if this means the end of the CDO role?

While we do not think Data Mesh means the end of the CDO, we do think that in organizations that adopt Data Mesh, the role of the CDO will change over time.

At the time of writing, we think there are a number of  reasons why organizations, even after implementing Data Mesh, will still need a senior data leader :

Data governance and Data Mesh

If we think about the federated computational governance model of Data Mesh, it clearly has many benefits, but the responsibility for corporate data governance is now shared across Domains. This potentially results in policy issues falling between the cracks. Also we are reminded of the old adage that “if everyone is responsible, then no one is responsible.” If we think about the misuse of data, a failure with respect to data governance could result in a systemic failure of the organization. 

A solution to this scenario is that we have a senior leadership role that has ultimate accountability and oversight of enterprise wide data governance and data usage across all of the domains. This could include enterprise wide rules such as those that pertain to GDPR compliance.

Enabling Domains to be more efficient and productive with data. 

Aspects such as data literacy and the definition and management of an enterprise wide knowledge graph are cross domain concerns, and there are also optimization opportunities that would be powerful in driving efficiencies across domains. 

A specific example would be that a business needs to present data attributes in a particular way, consistently across all domains. In this case, having a single organizational unit that creates a library of defined functions that the domains can leverage, would lead to tremendous efficiency savings. This organizational unit would need to be led by a senior leader.

The rise of the Chief Growth Officer role

There is a place for a senior leader who can look across all the domains to strategically synthesize new business opportunities through exploiting existing and future data assets. 

While domains might be looking at the next two quarters, the senior leader will look for future threats and opportunities over a longer time period. As such, we might see the CDO role morph into a “Chief Growth Officer” type role.

CDO to CGO

Dare we say, over time we will see the CDO role transform away from its current responsibilities to a focus on business efficiency and strategic innovation. We’d love to hear your thoughts on this.

 

Free eBook

Data Products For Dummies

Unlock the value in your data

Start Free with
Starburst Galaxy

Up to $500 in usage credits included

  • Query your data lake fast with Starburst's best-in-class MPP SQL query engine
  • Get up and running in less than 5 minutes
  • Easily deploy clusters in AWS, Azure and Google Cloud
For more deployment options:
Download Starburst Enterprise

Please fill in all required fields and ensure you are using a valid email address.

s