According to analysis published in the Harvard Business Review, the average tenure for a Chief Data Officer is between two and two-and-a-half years. Once I ascended to the role I didn’t even make it that far before needing time out, although I was arguably very successful. 

The same HBR article pointed out that the CIO title was once jokingly said to mean Career Is Over. I think there’s a parallel there with the CDO. The second you add a C to someone’s title, it draws a lot more attention and expectation. The buck stops at you, and appropriately so. The challenge with all new C roles, like the CDO role, is that expectations vary, keep changing, and often are hard to live up to. 

Building on shifting foundations 
Despite working hard up front to establish a shared purpose with my board members, I ran into challenges. Why? Because board members change and I didn’t pay enough attention to that. Perhaps I was just relieved that the one who’d said, “We don’t need a chief data officer, and that’s an end to it.” had left. Just remember that with each new board member, business strategy is slightly (and sometimes substantially) reset. Which can mean potential disruption and new expectations for all C-suite titles. 

An additional challenge is having to report to IT or technology. Not recommended. This has nothing to do with personalities and more to do with the fact that the CIO is responsible for delivering a service to the organization from a centralized place. The CDO, in my belief, serves a far more decentralized role. Most data and analytics work is done by people that are not even in the same function as the CDO. A data transformation should support everyone in the data workforce, from the heavy Excel analyst to a data engineer, to achieve more. 

Another way to look at it is that the CDO role is more about transformation or operational excellence than delivery. In practice that means a CIO responds to board requests with “Let me solve that for you,” whereas a CDO often responds with “Let me work with your teams to solve that challenge.”

If I made one mistake (one that is probably common to internal candidates), it was wanting the CDO role so much that I didn’t negotiate enough for it. And I don’t mean personal remuneration. I mean saying things like: 

  • I’m going to need X, Y, and Z to succeed 
  • Let’s define our priorities and how we’re going to say no to things that don’t fall on this list 
  • I need to report directly to the board to do this job well

No one can do everything 
As a CDO I had a clear remit for changing how the company worked with data, from developing data culture and skills to data product strategy and improving data management. I was fine with that broad remit, even though it excluded many areas like data science, risks and regulations, data ethics, and monetization. 

But scopes tend to creep and some requests are very hard to say no to, like persistent board requests or advice from consultants who have never held operational data roles. The challenge with data is that a lot of these requests can sound logical and may be very vogue in current thought leadership but require much more work than people understand and deliver little value. 

A previous example of this in the data world was building single versions of the truth with perfect data models ready to support any business need. Most companies have been burned by the types of projects that attempt to deliver everything and in doing so achieve almost nothing. 

Without going into the request I eventually agreed to, it is easy to list examples of work like this that are in vogue today: 

  • Migrating all the data to the cloud 
  • Cataloging or mapping all your data 
  • Building a perfect data stack or platform 

None of these initiatives deliver a single use case or release any value and yet they all require significant investment and keep many skilled people off data work that could be adding value. My advice would be to prioritize what will generate the greatest business value first. 

Another phenomenon I experienced was what I called post-COVID timing. In essence, that means people want everything, and they want it now. This is the opposite of strategy and more the equivalent of five-year-old’s football game where everyone on the field is chasing after the ball. To be successful as a data organization you have to triage and prioritize. No organization has the capacity to do everything that is possible with data. 

If you ever find yourself as a CDO 
Based on my experience, here are a few best practices to follow. 

  1. Establish a common understanding for the role. This is especially critical if you’re the company’s first CDO. They managed without you, so why do they need you now? That’s up to you to articulate.
  2. Build common purpose with the board. You can work with your board, which has permission to change your scope, or ignore them and risk being whipsawed by change. I’d move quickly to establish a rapport and a quarterly review of existing responsibilities. They need to engage with the scope of your work in order to trust your decisions. All this means that you should insist on reporting directly to the board. That’s true for any C-level executive, in my opinion. 
  3. Take an either/or, not a both/and mindset. When you set expectations with regular CEO and board check-ins, you also can establish the trade-offs between current responsibilities and new ones you might be asked to take on. For every new assignment you’re asked to add to an already full schedule, your response should be, “Great. Which existing project would you like me to kill?” 
  4. Know what they’re looking for. A large part of my leaving my role is based on having done a lot of what I wanted to achieve. The role simply changed to an operational CDO, whereas I’m more comfortable with transformation. Once we had laid the groundwork for the data function, spending three years looking at master data strategy for our bill of materials was not a good fit for me. 
  5. Have fun. Take the opportunity to enjoy your job. The CDO role has barely started, but it will become even more important now the AI era is starting. And AI is fueled by good data. So enjoy what you do. It’s worth the stresses and strains of the role.