The State of the CDO

Nate Nelson: Hey everybody and welcome back to the DataMasters podcast from Tamr. I'm Nate Nelson, sitting with Mark Marinelli, head of product at Tamr. Who's going to introduce the subject and guest of today's show. Mark, how are you?

‍

Mark Marinelli:I'm doing fine Nate, how about you?

‍

Nate Nelson: Good.

‍

Mark Marinelli: So in this episode, we're going to talk about how the CDO role has evolved from being centered around compliance and regulation to one that's focused on driving business revenue. That's one takeaway from a survey of 300 C-suite executives and financial institutions on the topic of data and digital transformation. So we're going to hear from the creator of that survey, Randy Bean, he's the founder and CEO of NewVantage Partners, which has been conducting that survey since 2012. NewVantage advices Fortune 1000 clients on using data to improve business operations. He'll talk about what makes for a successful CDO, where the CDO should sit in an organization and why data initiatives need to be tied to business goals.

‍

Nate Nelson: Here is Randy Bean. Welcome, Randy. Thanks for sitting with me. Your report contains a lot of varied data, but here we're going to focus on your findings regarding the role of the chief data officer. Can you speak to how the CDO role has been changing and becoming more defined in recent years?

‍

Randy Bean: Well, this is a subject that's near and dear to my heart, and I could speak for hours on this, but let me give you some of the highlights, both from a data perspective and anecdotally. In 2009, following the financial services' crisis, one of the partners of our firm was actually hired to be the first enterprise chief data officer of a talk for a bank and this case city group. So was very much a witness to the establishment of this role from the outset, and in the early 2000s, or I should say it was 2009 when this individual was appointed. So in the early 2010s, I wrote a number of articles, extolling the need of organizations to appoint a chief data officer and how this role could be critical in helping organizations become data-driven and get a handle on their data.

‍

Randy Bean: The reality has turned out to be much different than the initial expectations. It was probably naive on my part and naive on the industry to expect that any single individual or any single role could tackle all of the challenges that organizations face in terms of managing data both because of the proliferation of data and the coordination that's required for an organization to be effective.

‍

Nate Nelson: This doesn't sound like great news, by the way you're speaking about it. What does the data tell us how many companies are making space to a point CDOs?

‍

Randy Bean: In 2012, in our first survey, only 12% of organizations indicated that they had appointed a chief data officer, in our 2019 survey, 67.9% had indicated that they now had a chief data officers or sharp increase more than five fold during that period. But in 2020, there was a slight decline to 57.3%. And among the things that we're saying is that many organizations have become disillusioned what their expectations of the chief data officer role for a number of reasons. Among the largest banks that I've referred to, a number of them are on their fourth, fifth, or even their sixth iteration of the chief data officer over the course of the past decade. Some firms that we've work with, prominent firms, have abandoned or temporarily stacked away from the role of the chief data officer. A lot of it has had to do with the expectations of the role.

‍

Nate Nelson: Randy is painting a picture there of a little bit of a disenchantment maybe, with individual CDOs, where some company's having already filled the role multiple times and some just giving up. Why do you think successful CDOs are so hard to find?

‍

Mark Marinelli: Well, we're asking a lot from the CDO, because we're looking for someone to drive collaboration across two camps, that's ICA and business. We're still trying to figure out how to work together after 30 years, there's a gulf in skills, in spheres of influence, in agendas, that's not going to be closed overnight by anyone. The talent necessary to make that happen isn't common, you have to find someone who possesses the technical knowhow to go deep with IT and understand what they're doing and why they're doing it. But also the business acumen to find the most valuable uses for data across a never ending stream of projects. Finding both of those in one body is not as easy as you think.

‍

Nate Nelson: Yeah. But come on Mark. I mean, as you just mentioned, these two sides of that coin have been working together over many decades. So you're telling me that there aren't plenty of people with sufficient cross IT business experience out there, ready to do this kind of work?

‍

Mark Marinelli: Well, there's certainly been people working at the boundaries, they've been working, struggling to work together, but there's lots of CDOs I think who've either been the most business savvy member of IT or the most technical member from the business side. But only now after a lot of experimentation, are we seeing more clearly defined role and some devoted specialization aligned to the particular needs of the position. These are things like, how do I prioritize projects across multiple departments and IT resource pools? How do I have executive representation of this function? Those really go beyond the skills required to play on either team historically.

‍

Randy Bean: We've seen a shift in the expectations over the course of the past decade, initially, in the early 2010s, many organizations will hire the chief data officer largely for regulatory and compliance initiatives. What we call the defensive role in recent years, more and more organizations are aspiring to hire chief data officers to play a role in revenue generation, new product development, playing more of an offensive role. A number of the folks that initially were hired into the chief data officer role became frustrated that most of their time was being spent on defensive regulatory and compliance initiatives, as opposed to offensive revenue generation and business growth initiatives. Similarly, there was frustration on the part of organizations in terms of what were the skills that made an effective chief data officer. In our survey, and it's varied over the course of the past several years, but there's been a breakout in terms of the primary skills that organizations equate with chief data officer success.

‍

Randy Bean: So the first criteria is external change agent or an outsider and that's grown substantially. In 2018, it was roughly a third that identified this as the criteria and in 2020 it's roughly half. So that's somebody who comes from outside, brings a fresh perspective, can shake up the organization and operate in that fashion. In 2018, almost the equivalent numbers said that an effective chief data officer had to be a company veteran or an insider who knew how the organization operated, knew how to get things done, could be effective working within that organization. But that number has declined to roughly half of that, 15.8% in 2020. In other words, organizations seem to be concluding that the insider, even though they know how the organization works, they don't understand how to transform an organization in such a critical way as to become data driven requires.

‍

Randy Bean: There was also several other responses in terms of the successful chief data officer, in 2018, just over 10% thought, it should be a line of business executive who owns business results and nearly a quarter now believe it should be somebody who owns business results. So in other words, tying the CDO to revenue responsibility, again, that remains low for most organizations, but more and more organizations are aspiring to that. And then the two other criteria that were mentioned as criteria for a successful chief data office, one being a data scientist or annex analytics leader, but that seems to have declined in recent years. And other being a technology executive, which also has not increased in recent years. So there's no common formula for a successful chief data officer because the position's been nascent and new for most organizations, there's been no common job description. Many organizations previously, some of those responsibilities resided with the chief information officer.

‍

Randy Bean: In early incarnations of the CDO role, often the CDO reported into the chief information officer. What we're seeing now with many organizations is that the chief data office is now a peer of the CIO reporting into a chief operating officer or a head of technology and operations, or even into CFO's.

‍

Nate Nelson: One of the more interesting numbers I found in your survey is with regard to the question of whether CDOs should be given a seat on the executive boards of their companies. According to what I see here, whereas a couple of years back around half of respondents said that the CDO should sit on the board. Now that number is down to 38% and now half think that the CDO should instead be reporting to the executive committee, not themselves on it. Randy, how do you react to results like that?

‍

Randy Bean: It's been really difficult for CDOs to be successful in their role for all of the factors mentioned. And in the 2020 survey for the first time we asked this question very bluntly, how would you characterize the success of the CDO within your organization? And we gave three options, struggling with turnover and roughly 23% responded that that's where their organization was nascent and evolving, that was the majority nearly 50% and successful and established approximately 27%. So if you think about that it's just over a quarter of organizations after a decade, that really find that they've gained significant and meaningful traction in terms of making the CDO a successful role within their organization. So I think that that in part contributes to the response in terms of where the CDO should sit in the organization. I think that a number of organizations because of their level of disillusionment or frustration have moved from a point of view that the CDO should report into the executive committee, as opposed to sitting on the executive committee.

‍

Randy Bean: I think that that's tied directly to this inability of organizations to gain the type of traction that they initially expected in the CDO role. And so, as a result of that, they're defaulting to traditional management structures where they're asking the CDO to report into the head of technology and operations, the CFO, the chief operating officer, or another member of the executive committee. It's worth noting that for the past decade, I've been moderating, organizing and hosting the keynote panel each summer at the MIT chief data officer symposium. So each year I bring together a group of chief data officers from large financial services and healthcare organizations, just to give you an idea from year to year. In our event coming up this summer, I'll be moderating the panel with chief data offices from MasterCard, CVSL, Bristol Myers, Squibb, American express and Bank of China. Last year, my panel included chief data officers from Walmart, Citizens Bank, JPMorgan Chase and Cigna. And in the previous year, chief data offices from General Electric, Morgan Stanley and General Motors.

‍

Randy Bean: And it's the common experience of these organizations that all of these chief data offices are really struggling to be heard, fighting for a voice at the table, trying to define this and established as a formalized high level business function within the large corporate structure, but we're not there yet.

‍

Nate Nelson: Mark, in your experience, how much does the management structure, so that side of things, influence the success of a CDO?

‍

Mark Marinelli: Well, there's a phenomenon of accountability without authority at play in organizations where the CDO isn't truly a C level officer, who is able to marshal all of the resources necessary to make an impact. CDOs need to make tough technology decisions and adjudicate conflicts between IT and business and often these teams are vying for project prioritization. If companies are taking a wait and see approach to fully support a C-suite data function, no wonder these CDOs struggle to move the needle in the face of lots of institutional inertia and resource contention. Is all the more reason to ensure that a CDO gets some really valuable projects completed quickly so that they can build the capital that will gain them the necessary remit to be strategically effective.

‍

Nate Nelson: We've talked thus far about some of the lukewarm results of your survey maybe, but how about the positive stuff? So of all the organizations that you've peered into, Randy, what is it that makes the successful ones successful? How are they able to utilize their CDOs to the fullest?

‍

Randy Bean: Yep. There's three characteristics that I could offer from what I see as successful organizations, I would say they would be patient, in other words, there's no silver bullet, don't expect miracles overnight, this takes time. Secondly, would be persistent, this is a challenge, an issue that you need to work on and work on consistently and you need to sustain those efforts over time. Too many organizations make a major commitment to data initiatives, data investments, the CDO role, but tend to abandon them or to reduce the level of support they have within the organization because they don't see results achieved over night. So organizations like Capital One, or even to other degrees, American Express or MasterCard, didn't get to the point where they are today without persisting over many years. The third point I'd make here is a collaboration. We see the most successful organizations as those organizations that have built a strong collaboration between the technology functions and the line of business functions.

‍

Randy Bean: Those organizations, where the data initiatives are driven from a strong, with strong business sponsorship and tied to a specific business goals and business outcomes, we see those as being the most successful. On the other side of the spectrum will be those organizations that build the capability, they don't have a strong business sponsorship in terms of defining news cases. We often say, when we go into organizations, identify a first question or a second question or third question, and let's pull the data that's needed to answer that question. And by answering that question and by delivering a specific business outcome, an organization can start to demonstrate quick wins, which build credibility, which lead to momentum. And it's through that process that organizations achieve success. Organizations that attempt to do a big bang or attempt to go off and build something for long periods of time, but not connected to day to day results. We see those as the instances where the organizations are less successful and when there's that integration and collaboration between the business and data and technology functions, we see the highest degrees of success.

‍

Nate Nelson: So that was Randy Bean. Mark, what are your thoughts coming out of this?

‍

Mark Marinelli: Well, while there's a lot still being defined about the CDO role, who the CDO reports to, how success is measured, what the job actually entails. I think one point for Randy's survey is clear, data projects need to be tied to business initiatives and generating revenue. CDO role is less about defensive tasks, that Randy mentioned, around compliance and more about going on offense and helping the bottom line. I think that's worth keeping in mind whether you're an experienced CDO looking for different projects or just starting your career in data and analytics.

‍

Nate Nelson: All right, thanks a lot to Randy Bean for speaking with us and thank you, Mark.

‍

Mark Marinelli: Thanks Nate. Take care.

‍

Nate Nelson: This has been the DataMasters podcast from Tamr. Thanks to everybody who's listening.