With an average lifespan of just 30 months, the chief data officer (CDO) has proven to be one of the most difficult roles to hold within any enterprise, but it’s also become one of the most critical to the success of any business.

“It’s a really tough job,” Manta founder and CEO Tomas Kratky said of the CDO position. In order to tackle all of the challenges that come with the role, he said these leaders need to understand how to navigate and contextualize their data environments, as well as create cultural changes in the way people work with and understand data once it’s in front of them.

“Most of our customers have over ten thousand applications, and each application has potentially multiple data sources, with hundreds or thousands or even millions of data objects … or billions or trillions, it doesn't matter,” he added.

And a recent Gartner research shows many CDOs – or chief data and analytics officers (CDAO) – aren’t confident with their progress in handling the sheer volume of data within their organizations. Less than half of data and analytics leaders say their team is providing value to the enterprise.

Adding further complexity, many large enterprises have what Kratky calls a highly federated ownership of data sources and processes, where each part of an organization has a stake in a different subset of data.

“The CDO role was officially created because we felt that someone has to deal with all the diversity and make sure that even though we have such a federated environment, we can still have high-quality data and use it for daily decision-making,” he explained. “That is always a very hard task, unfortunately, and that's one of the key issues around CDOs.”

Federated data management is now being “born again” through trending technologies like data mesh infrastructure, but Kratky said he still sees many customers managing their data the opposite way.

“When you have someone who is thinking more like a technologist, what they want to do is build the infrastructure first, put all these pieces in place and then when it’s done, everything will be so easy and all these use cases will be so easy, but it's already too late and you are fired,” he said.

With data being dubbed by many as “the new oil,” visibility has been touted as an essential feature of data management architecture. But according to Kratky, perhaps a bit too much weight has been put on the concept of total visibility.

Complex infrastructure projects that provide visibility into the entirety of a company’s data environment take time and skill, but often these resources are being wasted on what Kratky likened to boiling the ocean. Instead, he said CDOs and data management teams should focus on context, honing in on select data sets that will actually deliver value.

“Really focus on powerful data use cases. Focus on the data set you need for that use case. Try to understand the whole business process and data flow for the data set and everyone who is involved and try to do something with this first,” he said.

This approach should apply if an enterprise has 10 critical data elements, ten thousand critical data elements, and even for those who just need visibility into a single data set.

“Having everything available doesn't help. Most people don't know how to start or where to start,” Kratky added. “You need to limit it to a really useful context and then deliver that context in a way you can actually use it and benefit from it. And that's the real magic here. How to do it right. Not how to do more.”

Misconceptions about the CDO role have led to shorter tenures than most of their C-level counterparts, with Kratky noting many enterprises hire CDOs without well-defined expectations for the role, and still expect those in the position to “change everything.”

“For some it's more about a digital transformation, needing to modernize. For some, it's more about compliance. It's more about risk and safety,” he said. “And for some, it's all of that and they don't communicate it well. They don't set expectations really well and then they let CDOs fail.”

For example, large enterprises that have over a hundred business units, each operating with different governance procedures and using and producing different data sets, creates a complex network of dependencies and relationships. The expectation for one person to come and fix data infrastructure is just not realistic.

“Even software engineers like myself are not trained to communicate about data. And now you hire someone and you expect them to fix it in a year or two. How is that even possible,” Kratky said. “It's hard in a small company like Manta, and think about a bank with 50,000 employees, it’s just crazy.”

Backing this up, Gartner claims that only 43% of top-performing leaders reported effectiveness in committing time to their own professional development, compared with 19% of low performers.

“Successful CDAOs must be elite leaders,” said Alan Duncan, distinguished VP analyst at Gartner, in a statement. “Top-performing CDAOs invest in their success by developing skills to thrive in ambiguous circumstances, articulate compelling value stories, and identify data and analytics products and services that can drive business impact.”

Also critical is the CDO making sure that employees “do not see data as something just sitting inside a company, because data is the company,” Kratky said. “I really believe that one of the key issues for everyone is not really understanding how data is everywhere.”

It's about the “processes and humans, and understanding how the organization is structured, and how to best use the structure to actually do more with data and do it better,” he added.

This puts a responsibility on the CDO to work with a lot of different parties, hopefully with the support of their CEO, to make changes that better align people with the context of the data they work with.

“One of the key issues that's not well-understood by CEOs and CIOs is how much this is about people and how it's actually not about technology,” Kratky said. “It's more about changing the way we think and how we operate and how people interact with data.”