Chief data officer: My mixed and nuanced musings on the need for one

Big Data Evangelist, IBM

Back in college, I recall my political science professor expressing amusement at the fact that some OPEC member nations have a cabinet-level "Chief Oil Minister."

After all, he said, that's just an export commodity. It's not a senior level position found in most governments (such as defense, treasury and foreign affairs), a key piece of national infrastructure (like transportation, communications and post office) or some set of programs critical to national well being (for example: agriculture, commerce, environmental protection, housing and education). It's not even the larger staple of a functioning economy (energy) that this particular commodity provides. It's just oil.

Oil rig 178581764.jpg

“What if,” my professor mused, “the USA had, say, a ‘Chief Corn Minister’?” But he was no dummy and he wasn't holding up a particular nation or ministry as a joke; rather, he used this rhetorical question as a springboard into a larger observation. He said that particular organizations dedicate particular executives to particular functions if those functions are sufficiently fundamental to the identity, survival or prosperity of the organization in question. For some countries whose economies and budgets depend largely on a single export product, it makes perfect sense to have a C-level individual for whom that's his or her sole focus.

When people say that "data is the new oil," they're usually making a general statement on how deeply modern organizations depend on data to drive transactions, analytics and processes in general. It's not a statement about public sector institutions but about organizations of any sort.

It's in that context that many organizations decide to appoint something called a chief data officer (CDO) to oversee this precious resource. If you want a deep dive into what the CDO role entails, I strongly urge you to download this excellent whitepaper from the IBM Center for Applied Insights. For your further perusal, here's a recent report on the IBM CDO Strategy Summit, authored by Shadi Copty, manager of strategy and operations in IBM Research's Almaden Lab.

Personally, I'm not sure that the responsibilities of a CDO, as described in these sources and elsewhere, are all that different from the older concept of a chief information officer (CIO): both of these positions, as usually defined, focus on leveraging data and analytics to drive better decision support and productivity at all levels of the organization, and both of them address data from the business level, leaving the operational, engineering and development side of it all to a chief technology officer, or equivalent.

Regardless of what we call this business-focused function (CDO, CIO or even chief analytics officer), it's important that there be a C-level executive who is responsible for the following:

  • Overseeing governance of the organization's data resources
  • Ensuring that they be applied effectively to achieve important business outcomes
  • Transforming the organization into a more data-savvy culture

Or is it?

In a data-savvy environment where data is integral to most business operations and where more business units can independently acquire the infrastructure, applications and data they need (through outsourcers, cloud services and more), the need for a C-level function such as CDO or CIO might strike some as superfluous.

I'm not the only one who thinks this. The above-cited whitepaper dissects the principal reasons why organizations might resist creating a CDO. These include:

  • The "territorial stance on data ownership" by business units
  • The "belief that the work was supported by existing rules and teams, such as the CIO and individual business units"
  • "Budgetary pressures"
  • "The difficulty in finding candidates who possessed both the technical and business skills required for the position"

Honestly, I think one could frame these points in a less defensive manner that goes to the heart of why a CDO or CIO might not make sense in some organizations.

Individual business units might appear "territorial" if what they're doing now for themselves works well enough--in other words, if they don't perceive significant value add in having to defer to extra special C-level oversight or give up control over their existing data.

The business units might appear to be citing "budgetary pressures" for this resistance when in fact they've already found a low-cost self-service alternative (such as cloud analytics) that fits their current budget.

And the seeming lack of suitable CDO candidates might be an illusion, due to the fact that business units' own executives might have all the requisite skills.

In that regard, here's a post I authored 2 years ago on the diversity of big data killer apps, which exist for every C-level function. As I stated there: "information technology, once the sole domain of the CIO and IT professionals, now features prominently on the agenda of virtually everyone in the C-suite, including the CEO." Everybody in the executive suite (from chief marketing officers to chief operations officers, chief information security officers, chief financial officers and beyond) probably has more data skills than most outsiders realize.

Why would any of them choose to accept a CDO position when what they're doing now puts them closer to the day-to-day excitement of using data to run the business?