Chief Analytics Officer & Chief Data Officer
Nothing says important like a "C" in your title. And it's been a long while since the C-suite had only room for the CEO, CFO, and COO. In the states, the CIO made its first appearance in the late 1980s or early '90s. Likewise the CMO. Now the C-suite literally bulges with chiefs. CTO, CHRO (human resources), CCO (compliance), CBO (brand), and so on. It's very crowded in there. And some C acronyms have to do double or even triple duty. For example, CCO can also stand for chief content officer or chief customer officer.
But as Thomas Davenport points out, organizations don't create new chief roles for no reason. On the contrary, when they do it indicates how seriously the organization takes a particular job function. Which brings us to a pair of relatively new C-suite entries: the chief analytics officer (CAO) and the chief data officer (CDO).
In early 2014, Gartner predicted that by 2015, 25 percent of large, global organizations would have a CDO. Do they? Who knows. But if Google search returns are any indication—and they usually are—more and more organizations are creating job descriptions and looking for CDOs and CAOs. Trying to figure out what each would do and how they would work together.
Competing on Analytics
What's driving this is Davenport's competing on analytics meme. Organizations that see their future success in data and analytics feel compelled to provide strong leadership in these areas.
CDOs and CAOs lead the effort to make their organizations more data driven. The terms “data driven” can be defined in a lot of ways, but for me they simply mean making smarter decisions based on analytics. Let’s take Capital One--the credit card company--as an example.
Capital One found its niche based on the philosophy of test and learn, with which it challenged the status quo idea that only your credit score could indicate a potential creditworthiness. The company believed there were creditworthy submarkets, and it pursued this idea in a very structured and analytical way. Its hypothesis was that in the aggregate there are creditworthy individuals who don't have great credit scores.
Every year, the company conducts tens of thousands of small sample size tests, which it analyzes in detail. It tests card features, risk models, offers, and advertising. And Capital One has been doing this since the 90s--long before anyone put the words “big”and “data” together.
That's probably an extreme example of how an organization strives to be data driven. But for CAOs and CDOs it reveals the two sides of a data-driven organization--the data and the analytics. And it also raises the question of how the two roles work together.
Supply and Demand
I tend to think of the CAO as being on the demand side of the data-driven organization while the CDO is on the supply side. It’s the CDO’s job to keep the kinks out of the data supply chain, to provide the data and data management.
And ensure that data assets can be used throughout an organization.
CDOs want their data assets to be discoverable. Yet they also want them to be secure. They don't want proprietary information inadvertently leaked. Of course, if the supply side can't meet demand, then the demand side finds it’s own supply. That’s where shadow IT comes in.
It’s important that the CDO and CAO share a vision of how data should be used in the organization. Developing that vision may mean the organization has to rethink the role of data and how it’s consumed. When data is plentiful and business changes at a rapid pace, the focus needs to be on data availability rather than data quality--at least for the demand side. And that requires more flexible policies on the supply side. Think of it as “loose coupling” between data access and data quality.
For this to work, the CDO must make clear that some data is “user beware” or “use at your own risk.” Likewise, the CAO must recognize this and take responsibility for how that data is used, who uses it, and the actions or decisions that flow from it.
This approach addresses one of the primary access versus quality issues for the CDO, which is being held accountable for data that hasn’t been vetted, normalized, cleansed, and so forth. It is an agreement that puts a premium on data exploration, which the demand side can use to understand what from the supply side may be useful for the business.
The loosely coupled relationship between supply and demand only works if the CAO and CDO work together to bring about the necessary changes to an organization’s data and analytics culture. You can kind of sum up the cultural change like this, “If someone analyzes a large set of very fresh data that has not been cleaned and misinterprets it leading to a bad business decision is that worse than someone using their gut and arriving at the same bad decision?”
My answer would be no. A bad decision is a bad decision. At least with the data-driven decision, there’s a place to start to determine what went wrong.
Of course this attitude pushes analytical decision making or at least analytical decision recommendation further down into the organization. To revisit Capital One for a moment, its recruitment process even at the junior level puts candidates through a series of analytical interviews. And that's not an unreasonable expectation within a data-driven company, where employees are expected to be critical thinkers that can be trusted to use data intelligently.
Nevertheless, for some companies this is a massive sea change. And the roles of CAO and CDO will be tasked with navigating it.