(I originally wrote this post for my good friends at Snowplow in early 2020. A year on, with another CDO position under my belt, I thought it would be timely to repost it here, lightly edited and updated.)
The 2010s were a big decade for Chief Data Officers: from a standing start ten years ago, CDO has risen to become an indispensable C-suite role, with almost two thirds of Fortune 500 organizations hiring one, according to a survey in 2019 by NewVantage Partners.
But the role of CDO, especially outside the US, is still poorly defined, and CDOs are frequently not set up for success within their organizations. Chris Bergh, CEO of DataKitchen, reckons that the average CDO tenure is just two and a half years (though this figure may be skewed by the newness of the role); clearly, CDOs are not being given much time to demonstrate their value before they are asked to move on, or (just as likely) find that the grass looks a little greener somewhere else.
Against this backdrop, should we conclude that the CDO role is a poisoned chalice? Should aspiring CDOs content themselves with less risky Head of Data or Head of Analytics roles, and wait for the role and its expectations to become a little clearer? Or should CDOs resign themselves to the fact that, to borrow from Hobbes, their lives will be nasty, brutish and short?
What even is a CDO, anyway?
One of the biggest challenges with the CDO role is that its definition is poorly agreed across the industry, and is changing rapidly. According to Mario Faria of Gartner, we’re already onto Version 4.0 of the CDO role. Since most organizations have barely come to grips with versions 1 to 3 of a CDO, it’s easy to understand why there is confusion.
My own definition of the CDO role encompasses the following broad areas of responsibility:
- Building and managing a robust, broad-scale data platform (likely in partnership with CTO)
- Directing a team of analysts and data scientists to deliver insights from the data
- Creating an operational model for data utilization across the organization
- Engaging with senior stakeholders to solicit data needs and provide inspiration/ideation for the use of data to support company objectives
- Managing data governance and privacy
- Driving data-driven/data-supported behaviours and cultural change, as part of broader company-wide digital transformation initiatives
As important as the areas of CDO responsibility, however, is the way that the CDO operates. A great deal of data and analytics work, even at large, sophisticated companies, is done on a highly ad hoc basis, with little of the management rigor that has become commonplace in the world of software development. This causes two problems:
- First, stakeholders find it very hard to understand whether (and when) their requirements will be met by the data team, encouraging them to go off and roll their own solutions to their business questions, often with disastrous results.
- Second, the individuals on the data team burn out because they always feel overwhelmed with work, and have no method for prioritizing the many conflicting demands on their time. This, of course, causes attrition, which further exacerbates the problem.
The most effective CDOs act as the “glue” between the business and the data team, providing rigorous oversight of the work the data team is doing, while skilfully managing the expectations of the stakeholders who are dependent on that work. To do this, the CDO must exhibit breadth of ability across the areas above, and be able to stitch them together to deliver a compelling data capability for the organization. It’s not an easy job, but it’s made considerably more difficult by a number of ways that organizations (especially their senior leadership) can undermine the CDO role.
How to make your CDO’s life a misery
Here are just a few of the ways that well-meaning organizations can make their CDO’s life much more difficult than it needs to be.
Hire them before you figure out what their job is
Hiring anyone before you’ve decided what their goals and objectives are, and how their success will be measured, is a terrible idea, even in a senior role like CDO. If you can’t figure out what impact you’d like your CDO to have on your business, you probably shouldn’t be hiring one yet.
The counterpoint to this argument runs, “but we hired the CDO to figure out what we should be doing with data, and then we’ll implement their recommendations.” This sounds good, but is tremendously risky. If you cannot implement the CDO’s recommendations in a timely fashion, they’ll likely leave, putting you back to square one; and even the most enlightened and even-handed CDO will want to build out the function in a way that plays to their strengths and interests. Much better instead to hire a consultant to develop an overarching data strategy, including a recommendation on the CDO role, responsibilities and profile you need, and then ensure you have resources in place before you hire against that profile.
Don’t empower them to do anything
Almost worse than hiring a CDO without knowing what you expect of them is to hire someone and give them a set of goals without empowering them to reach those goals. For CDOs, this issue most often arises when an organization’s data assets (both data and people) are decentralized; the first job of the CDO then becomes to persuade a whole bunch of people to give up their own resources and hand them over to the data team. Without the ability to show some benefit of centralization of resources (which is very hard to demonstrate before the resources have been centralized) this becomes a very long and boring game of chicken and egg. A whole year can go by just trying to solve this problem – and if it takes a year, it will probably be the only year of your CDO’s tenure, as they’ll leave for somewhere where they actually have some power. Don’t make your CDO a paper tiger.
Ignore where you are on your data journey
The CDO you need to get you off the blocks with a basic set of usable data, accessible by a small group of specialists, is not necessarily the CDO you need to scale your data capabilities to thousands of individuals across your organization and drive broad-scale culture change. As companies are building out their core data foundations, a lot of core technical engineering work is required to gather (or centralize) data, clean it up, and store it somewhere useful; the right leader for this is someone who can drive this kind of detailed work hard for 12 – 24 months without getting too distracted by day-to-day considerations.
Once the platform is in place, though, you need someone who can take that data and use it to inspire conversations up and down the company about the possibilities that the data unlocks, and bring along sometimes skeptical stakeholders (many of whom will be in the C-suite) on the journey. This is likely to be a very different person from the one who built the platform (though a few CDOs can do both).
Overwhelm them with “business as usual”
A counterpart to the “paper tiger” CDO who doesn’t really have a team is the one who is hired to manage an existing BI team that has been cranking out the same old financial reports for years. All efforts by the CDO to dial back the laborious production of these existing reports are rebuffed, since they’re all ‘essential’. This prevents the CDO from freeing up precious resources to focus on new projects that will deliver incremental value.
Define their role too narrowly
Many companies hire a CDO because they have a specific, narrow problem to solve, such as data governance (many CDO roles were created in 2018 to deal with GDPR), or data migration/connection, often having the role report into the CTO, or even the finance or legal teams. The trouble is, once the individuals in these roles have solved the immediate problem they were hired to fix, they struggle to deliver broader value across the organization, because they don’t have the mandate to do so.
Expect miracles, instantly
Even if you have empowered your CDO to drive real change, and have broad organizational buy-in for these changes, it will take them some time to deliver significant value, especially if your data is incomplete, of poor quality, or fragmented across your organization (and if it weren’t, you likely wouldn’t be hiring a CDO in the first place). Successful CDOs do need to act with a strong sense of urgency, and find ways to deliver incremental value, but expecting a rapid transformation of the company’s fortunes over a short period of time is unrealistic.
Insist they only save money
Of course, data and the automation that relies upon it can drive significant efficiencies within an organization. But any effective CDO should also be looking for ways to drive top-line revenue growth for the company, perhaps by increasing sales directly through more effective marketing, or by driving product improvements. It is likely, however, that in order to realize these benefits, the CDO may need to spend some money. Don’t cripple them by insisting their entire operation be “self-funding”, or treating them purely as a cost centre.
The good news: Life for the CDO is improving
Fortunately, organizations are becoming somewhat more enlightened about their CDOs, and seem to be taking the role more seriously, partly as a result of the central role that CDOs can play in digital transformation, which has been brought to the fore by COVID-19 and the challenges it has thrown up.
In the latest (2021) iteration of their annual Big Data & AI Survey, NewVantage Partners asked a group of Fortune 500 Finance and Healthcare companies about their CDOs and their efforts around data more broadly.
Here’s what these companies want from their CDO:
- 40% want an external change agent for their CDO role (down from 44% in 2020), versus 23% wanting an insider (up from 16%)
- 50% (up from 40%) want the CDO to have overall data ownership for the organization
- 47% (up from 38%) see the CDO as a tier-1 (board-reporting) role, with only 4% seeing the role as interim or unnecessary.
- Most interesting of all, fully 92% of respondents cited people, process or culture as the main impediments to progress in data for their organizations, with only 8% stating that technology was the main blocker.
Clearly the CDO role has a long way to go before it is a well-established (and well-understood) leadership role within many organizations. However, companies are beginning to see that the role is essential not just in putting in place the right technology and foundations for data-driven transformation, but also to lead the cultural and organizational change that is necessary for them to undergo their digital transformations.