Who Reports to Who? Asking for a CDO

0:01 Welcome to The Prompt, our short-format minisode from How I Met Your Data, where Junaid, Karen, and I unpack the latest in data and AI with sharp takes and real-time reactions. Think of it as our hot take session on the topics making waves, from industry shakeups to emerging trends, all served with our usual curiosity and candor. Quick disclaimer, these are our personal views, not the views of our employers. So, grab a coffee, get comfy, and let's dive in. Junaid, Karen, I'm channeling my inner Linda Richmond right now with a 1990s SNL throwback.

0:42 I feel like we need a quick coffee talk. I'll give you a topic and discuss amongst yourselves. The CDO or CDAO, who should they report to and why? So I'll kick it off. I have over the course of my career in data, I've worked up to I think every C-level person. I've worked up through a COO, a CFO, a CRO, and there's cases for each. So if you're working on reg oriented stuff like finance heavy things maybe this the cfo that should be accountable for the financial data right if you're if you're all about your public financial data what you're giving certain regulators maybe that makes sense if there are operational issues that rely heavily on data maybe it's the coo another argument can be made if you're a lot of reg work for the chief risk officer because of the nature of data and the risk that it presents.

1:40 Data is an input to strategic decisions in the boardroom. And so you want to mitigate risk that happens there. I'm still a little bit old school. And I would say a long time ago, I would say, or maybe when this role first started appearing, it was reporting up to CIOs and CTO, which are another option, I think, to really enable data, drive data to be leveraged as an asset in a firm. I think it's still got to go to a CEO. That would be my, in an ideal world, a CEO should have the same literacy that he has on financial data.

2:21 Like a CEO can talk about earnings per share and talk about EBITDA and talk about all of these things. And I think the future relies on CEOs being just as literate on data and the impacts of data and AI, right? and how AI will come into the picture. So I think if I had one person, I think it would go to the CEO. If it's not going to go to the CEO, I would go, let's have the role report to a CIO, which would then, the CIO would have both the chief data officer reporting to him and the chief technology officer reporting to him.

3:02 So the CDO and the CTO would be peers in that scenario. That's my perspective. We'd love to hear what you guys think. Wow. I mean, I agree with your outcome, but how you get there, I think I feel like there's a lot of subtext that we're not talking about. For example, there's the parody of your remits, what you're being asked to do and what you're being judged on to be successful and the agency you have to do it. And I think the subtext of what we're seeing or why we're even talking about it is because I don't know the specific figure these days, but it's known that the tenure of most.

3:45 CD, fill in the blank for whatever else comes after the acronym, is that it's short-lived. It's usually about 18 months. And often the hiring cycle for that role, regardless of who you report to. Focuses on skills that are not necessary for you to be successful, meaning that you have to, as a leader, understand the connection to the P&L and how data relates to that. However, often the job descriptions focus on exceptionally technical skills, which are helpful, but that's not really what you're going to be doing in the day-to-day. And so often what happens is that we see that someone is they go through the hiring cycle they're exactly what they put in the job description and then they are let go because they didn't deliver on the strategic goals that were given to them and then you're like well what did you expect right because you didn't prioritize the actual skill that said that you need to be successful so in that respect I think there's a decision of what is actually needed for the role and I think that's more the subtext of what's not discussed.

4:53 And then further, is data something that's a cost center or an opportunity? And so I think if you recognize it as being something that is an asset and that is a revenue opportunity, then absolutely, I feel all day long it should report to the CEO. However, I feel like when it's a cost center, that's when you get relegated to reporting to someone else. And that's really a tell from an organization often what's in the job description and who you're reporting to as to what you should expect from the role and red flags along the way. But, I mean, that's the whole thing. In this day and age, is it even realistic or viable for an organization to view their data as a cost center as opposed to a strategic asset?

5:42 Right. No, I mean, I think the answer should be no. And I would say, you know, I would say this, that, you know, you touched on a very good topic, which is why is the CDO position or tenure so short? You're right, between 18 and 36 months since the vast majority of 10 years for CDOs. I think it's because there's still not enough data literacy in the C-suite. There's not enough investment. I think there's three components for a CDO to be successful. They need investment. So investment in technology, investment in resources. So they need budget. They need time.

6:25 Changing the data landscape is like steering a battleship. Nothing happens to us. I don't care what institution you are, how agile you are. You think you are, but you need time to change the landscape. And in a C-suite position, the data, the chief data officer, the chief data AI officer isn't afforded the same time as CEO is, which is typically five, six, seven years, I think is the average. And then the third thing that you need is influence to be successful as the chief data and AI officer. And that's why I had said let's have the chief data and the AI officer report to the CEO because at that level they will get influence because they'll be in the sort of the operating committee or that first line or like CEO minus one they'll have access to to budget and hopefully in a position to educate his c-suite peers on data to afford himself more time and I think you know again I think you know when you say it that way, it's a good question. Why do CDOs have this short tenure?

7:33 Well, if they and then kind of bridging it to the three things that they need to succeed makes an even stronger case for them to report to the to the CEO, I think. And I think the tenure question is actually a much longer conversation, but I was fortunate enough last year to present at the CDAO symposium in Boston with Malcolm Hawker on the topic of data culture. And we started the conversation with why is the tenure of the CDO so short? And Gartner points to CDOs saying that the organization that they were brought into doesn't have a data culture.

8:17 And we really pressed on that and really poked at it saying, actually, that's not true. Organizations have culture. Culture exists amongst people. So if you're working for an organization, there is a culture there. Guess what? It's just not the culture that you want or that you expected. I would argue that's a bit of a cop-out. 100%. To say that they didn't have a data culture, right? It just wasn't the culture that you wanted. But you do hear that often, which, I mean, again, I like to think of the subtext of like, yes, maybe they didn't have a daily culture. They did have a culture, just not one that you were able to like airdrop into to deliver amazing data products.

9:03 Within a few months. I mean, that's just like, not realistic. I might take an opposing view to this concept of a data culture versus it being a thing or not. I think there is such a thing as data culture, right? When I think about data, data 20, 30, 40 years ago was a tech function. It sat inside of technology. So whoever was deploying software, whoever was delivering technology embedded in it was they were delivering the data around it. I don't think 20, 25 years ago that they thought about data lineage in the way that we think about it today.

9:47 25 years ago, and I can say that because I grew up in technology 25 years ago in a tech function, delivering technology. And you would, when you think about data lineage, you would think about the single hop. Where does the data come from? And that would be basically it. You never went to the source. You just knew you got it from this place, which was just one hop away. And I think that what's happened in the last 15, 20 years, data as a discipline, since it's been sort of pulled out of technology, is looked at with this nuance of, hey, data has this life cycle that is separate from the technology life cycle.

10:28 And you know the tech life cycle is very specific to delivering technology and hardware and software the data life cycle is something that is very different in terms of how data is created the journey that data takes through it and the infrastructure of of a company and i'll say this data is separate than technology because the data as an asset over time even though data decays becomes more valuable than the technology itself. So like the data that is created over time, data that is, even though it may become dated, has value in terms of its historical significance.

11:14 And that data eventually outlives the software that it sits in. And I think that the notion of a data culture is people, you know. I've gone into places where they still see data as a tech thing. Oh, the guys in the basement working on technology are doing this. Whereas the nuance as it relates to data quality and its dimensions like I said in the late 90s early 2000s we didn't have this concept of completeness and validity and actually like the, nuance that we're looking at at it is so different and I think that's for me when I hear the word data culture it's having that ability to look at that look at us in that lens of nuance there's so much to unpack there I mean I would also challenge what you're I agree in the sense that in order to get the data that really means something to you, the technology has to enable it and the people have to participate in the process.

12:14 All of it has to come together. And ultimately, it doesn't matter how great your technology is, the people are always a point of failure. And I would argue that why does anyone else in the organization care? How are they incentivized to care about data in the sense that if it's serving their own needs to help them make their goals, then they will naturally become more interested in data and making it right. If the bar is low for what you're asking of them and they can get something out of it. I mean, it's basic human nature and it kind of goes back to my concept of data therapy and that, you know. It's the human psychology of all of it that really has to work for everyone, for everything to click to get the value from the data. And that, I think there are a lot of cheesy metaphors where everyone talks about, is Jada the gold?

13:04 Is it the blood of the, you know, pulsing through the organization? And like something that clicked with me that I was like, oh, hold on to that. It's like uranium in that used properly, it can power the whole organization. However, if not handled well, it can become exceptionally toxic to everyone around and causes more problems that you have to solve and that really resonated more of like what's possible but the reality of what happens where everyone's trying to serve their own needs and often have competing goals and deliverables where it's not in their best interest to opt into the data culture so do you think like to to propel a data culture where do you think this the cdea cdo or cdeio should report to i think that makes another case for the ceo is like if you want if you think that if you think that data culture is a thing right let me accept that then you need to like okay you need you need the cdo cdio in direct to a ceo if you don't think data culture is a thing then it becomes less important where where the cda cdo cdio reports to i mean i think there's like a chicken and egg conversation here.

14:20 And that accepting when you walk in the door that it's very difficult, complex, no one's going to get it. No one will have an appetite on top of everything else they have during the day to opt into your initiatives. Why should they do one more thing to help you meet your needs or your goals when you are brought on board? And that it really is a long game. And so how do you overcome things like, you know, lip service, showing up to the meetings and then zero follow-through. These are all known, and it's very difficult when you're coming in the front door to quantify ROI, measuring results that you can champion.

15:01 And so if you're starting off with unrealistic expectations, and to your point, Jenae, have no investment, does it matter? I would love to hear your guys' final vote on this. My final vote would be, I think, where it started. Directing the CEO, if that's not possible, to a chief informational officer that manages both technology and data so that tech and data can be pairs. Though I would say this, it can work, it can, I will qualify this to say, saying that I think the position can work in a COO, can work for the CRO, can work for others in the C-suite, depending on the firm's needs.

15:41 Yes, I do agree that ideally the CDO, CDAO role should report to the CEO or alternatively the CIO to bring equal strategic and technical attention to both data and technology. I am struggling a bit to reconcile the idea of positioning a strategic data role within, say, a regulatory or finance function, because I feel functional level alignment dilutes the ability to manage data strategically at the enterprise level. It's going to be interesting to see how this develops in the next, now, two to five years.

16:25 I think as, as people in general become more data literate, it'd be interesting to see if the CDO tenures get longer, right? It'll be interesting for sure. There's consensus that is the CEO role should be, if you value data as an asset, where the role of report, I think my, I don't have, I'll throw in. And if it doesn't have a clear, we all don't have a clear understanding of what a CDO is and how it's supposed to work in an organization, because there are so many definitions, it's just such a meaningless word, because it has no relation to the role, because it means everything, it means nothing.

17:12 And that's really where you get lost, even opening up the conversation, I think. So final question on the topic given the relatively short tenure of the cdo cdao role hire from the outside or promote from within and why that's a that's a tough one i'm gonna go when you need a change agent or someone and someone who's been there and done it needs to either change the culture and bring in expertise, then you've got to hire externally. You needed someone that isn't going to. You know uh you're gonna do something like more of the same sort of saying you need you know a change agent has to be generally someone from external from outside the the organization where, there is a where there is a data culture there is an emphasis on the role there is an understanding of what the remit should be and you then you'd want to hire internally that person will know the mechanics of the organization and how to get things done.

18:18 And then I think if you're external, if you're a CDO external hire, then you want your team, you want to hire most of your team internally because then they can compensate for the institutional knowledge. And then if you are an internal hire, then you want to hire people who are external, who've done this work somewhere else. So that would be my guidance. It's such a spicy question. I mean, yes, I think it comes back to more understanding. Change agent can come from many places. And I do think I have bias towards hiring internally for the reason that you are already, in theory, understand how the business works, have relationships with folks, which is really what you're going to be doing is managing relationships.

19:04 Selling what you're trying to do to everyone else in the organization, and that everyone trusts you. And I think that that's a big concept that really is foundational to being successful. I would then add that a change agent can come from many different places. And I think that's where most people get hung up on. And that you can find a change agent that may be someone that's really killing it from the finance team, perhaps. It can come from like product owner perspective, although arguably they're already kind of tuned into data if they're doing products in some capacity. Or not, but that understands the connection of the data and how it adds value to the organization or once you're able to see kind of under the covers what's going on.

19:50 And that's more where I would argue a change agent can come from many places, but an internal hire, I think, would be optimal. But the reasons why someone would not think of that, I think, are a bigger tell of where they have red flags or things they need to work on for whomever takes on that role for them to be successful. Well, excellent both. Love the topic. Thank you for your insights. And we will do this again really soon.