Podcast: Understanding the role of the Chief Data OfficerVincent Yates
Credera is excited to announce the release of our latest podcast: "Understanding the role of the Chief Data Officer."
This podcast, which is available on iTunes, Spotify, Google, and Anchor FM, brings together some of the brightest sparks in technology and transformation consulting to wax lyrical on current trends and challenges facing organisations today.
On this episode
While Chief Data Officers have the privilege of managing the world’s most valuable asset, they’re in for a tough task. Most Chief Data Officers still struggle with the ambiguity of their role and the massive responsibility of transforming their organisations to become data-driven.
In our latest podcast, Chief Data Scientist Vincent Yates and Credera's CTO Jason Goth discuss a few key questions about the role of the CDO:
- What are the blockers to a CDOs success? How they can navigate those to find success?
- Who should CDOs report to?
- What does the role of the CDO need to include?
- If organisations don’t have a CDO, should hire internally or externally?
- What is the best way for CDOs to drive value?
The following transcript has been edited for clarity and length.
Vincent Yates (00:02):
Welcome to Technically Minded, a podcast brought to you by Credera. We get technology leaders together to discuss what's happening in our world. Our discussions are always fun, lighthearted, and frankly opinionated, but hopefully it gives you a sense of what matters, what to pay attention to, and what to ignore. Welcome Jason. Today I have a fun one for us.
Jason Goth (00:23):
All right. I'm looking forward to it.
Vincent Yates (00:25):
I want to talk about the role of a CDO. By the way, do you know what CDO means, Sarah?
Jason Goth (00:31):
Chief data officer or chief digital officer?
Sarah Barber (00:32):
I would've guessed both, but I did at least know.
Understanding the role of the Chief Data Officer
Vincent Yates (00:34):
All right, good work. That is correct. This is one of those overloaded terms. Yes. So chief data officer specifically, or I did a little bit of research before the podcast, the other common ones would be, of course, my title, chief data scientist, chief data and analytics officer, or just chief analytics officer. I spent, as you guys know, I spent last week over at MIT at the CDO, chief data officer, IQ conference up in Boston at MIT, and it was a really fun conference, it was amazing. A ton of CDOs were there, ranging from BestBuy to Scotia Bank to to a huge number of others as well. It was easily the most CDOs I had spent a significant amount of time with at once.
It was quite cool, really great event, but coming out of that, I mean, what I realised is that this role, while reasonably broad, as I just said, all those companies have one is incredibly massive and equally ambiguous. I mean, data crosses, if you think about a chief data officer, their job must be data. And if you think about data, it crosses the entire organisation and touches every business, every department, every executive that's already there. And it leaves these CDOs with this massive, massive area of responsibility. And what's remarkable about that is that despite all of that, and despite managing the, you know, the world's most valuable asset CEOs, the exact, the actual executive or board of directors, rarely if ever, I only found one example where they did this, but rarely if ever provide any clear objectives of what success looks like for that role beyond "make us more data driven".
And it seems to me like without a clear charter, without the accompanying authority, the process to transform an entire organisation's culture to be data-driven is at best, difficult, and at worst, futile. And thinking about that, I think there's a lot that goes into that. And I want to unpack that today.
But I was trying to figure out, are there places in the world that we've seen something like this happen before these transitions of this new, this new data in this case, but some other technology where we're going to introduce a new role and it's kind of nebulous. It's very broad. It's very difficult to get our arms around. And it occurred to me, Jason, that your role, our chief technology officer went through the same transition not so long ago.
Learning from the role of the Chief Information Officer
Jason Goth (03:05):
That's right. I think CIO or CTO, it depends on what companies call the role...
Vincent Yates (03:13):
By the way, I heard, I heard a great one real quick, sorry to interrupt you. CIO meant "career is over".
Jason Goth (03:18):
That's what it used to mean. Yes, that's right. But you know, companies have roles that are focused on internal technology, things like email and finance and, keeping the lights running. And then they work increasingly more to expose the company externally throughout the digital ecosystem. And some folks will call that role CIO; the internal one, CIO, and the external one CTO. Some people will call it the opposite. It doesn't really matter, you know, it depends on what you call it, but those are, are two pretty well defined roles.
Vincent Yates (03:50):
Okay. And are they typically distinct, by the way? Cause I wanna get into that with data versus data and analytics. I wanna talk about that too. Are they separate or are they the same usually?
Jason Goth (03:57):
They are typically separate roles for different people. Although for both, there may be a CIO that has reports responsible for both internal and external. Right. Again, that's very company / org structure specific. But the two roles, internal and external, do exist. And, and if you think about the external technology, it has certainly grown significantly in the last 20 years. I love to tell the story - there was an article back in 2003 in the Harvard Business Review by a guy named Nicholas Cars. I'm sure a lot of the listeners will have heard of it. It was an article that was called "IT doesn't matter".
Vincent Yates (04:38):
This was 2003?
Jason Goth (04:39):
Yes 2003, and his argument was that eventually, technology would all become commodities. Just like essentially like the telephone system. Telephone, fax, copier, they're all commodities. We go out, we hire AT&T to come and put phones on the desk. We don't like him. Next week we get Cisco and they replace it, you know, with their system voicemail and all of that. And then essentially all technology would become commoditised like that. Clearly, that didn't happen. Right? If you, if you look at what has happened since 2003 world has gotten more and more dependent on technology companies, whether you call it digital transformation or whatever. I love the Mark Anderson quote, "software's eating the world". Everything is driven by software, right? What do you do every day from your watch, looking at your watch to driving your car, to making an airline booking, to getting reservations for dinner tonight with my daughter, everything is driven by software and people expect more and more and more from the technology ecosystem.
Right. I expect to be able to order on Uber eats from any restaurant I want. So, companies have to continue to keep growing and adding these digital capabilities. And in a lot of ways, the CIO, I'll call the role CIO for the purpose of the podcast, but they have become very important to the long term success and strategy and survival of the organisation. And so it, it has gone from being, "we manage some internal systems like email and fax and finance" (that a small subset of internal users use and we need to be able to run the business, but it's, you know, somewhat commodity and, and a cost centre) to having to reimagine the role as someone that is really a business leader that works with the business to see how we can use technology to achieve those objectives.
And they had to reimagine how the entire software and technology delivery function works within the organisation. And they had to reevaluate how they look at talent and people and culture and everything to transition to these more technology focused organisations. And, and so that has been a big transition. And so where they are now, I don't know that it is perfectly defined, but I, I think it is a lot more defined. And the point of all this longwinded rambling answer is to say that I think the chief data officers are somewhere in that initial state where, it was data and we needed it and now it's becoming critical to the success of the business and what those different roles and responsibilities are is still somewhat nebulous, unclear or different in different companies.
The current state of CDOs
Vincent Yates (07:40):
Yeah. I think that's exactly right. And, and I'll just give you some stats here just to set the stage a bit more. There's a couple of stories I'm going to be referencing today. Let me just call them out ahead of time that way I don't have to keep pulling them up. The first is from new vantage partners. This is called the data and AI Leadership Executive Survey from 2022. Really good, lots of great data in there. The second one I'm gonna be referencing a fair amount is from PWC Strategy&, and they call it their 'In the Age of Data, Why Are There So Few Chief Data Officers'. Couple of things I'm going to call out right away. By the way, these do have different sample sizes in different respondents here. So there some stats aren't exactly compatible, but I'll do my best to clarify where that's appropriate.
The headline here is that out of the top 2,500 publicly traded companies worldwide, only 21% have a chief data officer at all. And most of those data officers have been hired in the past year or two. If you look at those roles, there isn't a consensus here yet, but most do not report to the CEO currently. And I want to talk about that some and how that played out for CIOs and CTOs historically. And the other interesting thing is that out of all of those people who have been appointed, currently the average tenure of a CDO is only 18 months. And so I think a lot of that has parallels to what you described that the CIOs and CTOs went through before, which is, again, you're responsible for this cost centre, The CDO largely role came out of compliance and regulatory requirements. If you look at the places where these things have better adoption, it's typically, insurance, it's typically banking, pharma, for example, places where you had some regulatory requirements around it. And they were very defensive in the way they talked about it. How did that play out for CIOs? Again, it must have been similar in that, hey, there's this eCommerce thing we have to get on the bandwagon. It feels a bit defensive. We don't do our competitor gonna do it. Maybe there's some innovation in there originally. But how did that play out for you all? And what could we learn from the data side?
Jason Goth (09:40):
I think CIOs before 2000 probably mostly reported to the CFO. Or if they did report to the CEO, they were not viewed as a strategic leader. They were viewed as someone that oversaw a very large cost centre that didn't have a lot of strategic bearing on the direction of the company. And it's really more cost of doing business. We have to have printers, we have to have copiers. We have to have phones. We have to have all these computer systems. I do think you called out one thing - the advent of the web and e-commerce really was a big driver. It's like, wait a minute. We have to now put these systems online. When we put them online, they have to scale, not to a hundred people in our call centre, but to potentially a hundred thousand people every day online.
The other aspect of that is security. When we put them online, we have a huge security issue. That data can be hacked leave and we can have a lot of bad outcomes. So that I think was one big turning point. The second turning point was in 2007 was the invention of the smartphone. And we put more and more capability data on smart devices and the people interact with every single day. How, how long do you go without using some, even getting your email on your phone, texting. Like, I don't want to tell you how much stuff I buy off Amazon prime on my phone, you know, tracking your heart rate. Your health and everything else, you know, purchasing airline tickets, changing your flights. I mean, how integral is your phone? And so that went from, well now not only do we have to publish this stuff out on the web, we have to publish this stuff out to mobile devices, which is a different form factor.
In those days, 2007, they had very little memory. Things had to be very scaled down. We couldn't really use the same solutions for both. And then a few years ago, the idea of the kind of digital ecosystem arose where it wasn't so much that you had to offer your products and services through the web or through mobile, but through an entire digital ecosystem. The example there is Uber Eats. Like, if I'm a restaurant, I don't just do online reservations. Now I take online orders through my own site or my own mobile app. I now have to take them through Uber Eats and DoorDash and plug in with all of these other ordering or delivery providers and start having to integrate externally. And most integrations today are through external systems through some type of API, not internal systems, and the way we started building systems relied on moving things more into this kind of digital ecosystem. An example there is payment. No one builds their own payment provider. They use one of the third parties that are available. And so these systems have become a very distributed network across both internal and external parties. And so as we do that, technology becomes more and more crucial to the functioning of the company. And, you know, that's what I think has led to the CIOs becoming more and more important in the company's leadership.
Vincent Yates (13:03):
Yeah. I think that makes sense. And again, reporting to the CFO sounds to me like what you're saying is at least early days before e-commerce, and this really drove a lot of business innovation and, and revenue, it was really cost savings. It was, how do we manage this cost? We know we have to do, it a bit defensive, and then just sort of putting a cap on the cost of it, is that right?
Jason Goth (13:20):
Vincent Yates (13:21):
And I think that aligns with what we see a lot from the surveys, at least about chief data officers or chief data analytics officers. If you look at the survey from NVP, it turns out that they say roughly 40 ish % are about defence, regulatory and efficiency savings. Whereas it does seem like there was a big uptick in 2021, where about 70% said it was offence. Now it's back down to about 64%. There's this tension that you described that CTOs and CIOs went through that they're feeling as well.
The other thing I wanted to point out is if you think about the actual value that the data's supposed to be creating, we all talk about it's the most valuable asset in the world. What's remarkable is that, and again, this is in NVP's survey, so this is again biased towards organisations that have data officers only, 53% have a corporate data strategy. Only half of those, that is half the companies out there out of a survey that has chief data officers, 74% of these people have chief data officers or somebody responsible for it, only half of those companies have a data strategy, corporate data strategy. And out of those only, only 60% have a strategy that's generating success.
So what we're really talking about here is a relatively new role that's permeating a fair decent percentage of the market, that's not too bad (of the top large companies anyway). But of those, only about a quarter are actually generating any real success, at least with their data strategy holistically. I think part of that is that they haven't been in the C-suite. They haven't been, they haven't had leadership roles historically. It's partly that a lot of it came out of that insurance and banking, where there was religious compliance and governance that they had to do for regulatory reasons. It wasn't seen as a big driver of the business, but we're seeing that change now where everything was AI just a couple years ago, everything was innovation. And now we're getting into some balance. Did that happen to CIO/CTO too, which is like, it started in CFO, it was cost control. It moved to the eCommerce and really growth functioning of the business. Did it move back? What's the balance there?
Jason Goth (15:27):
I think for the most part, most CIOs report to the CEO and are considered a key strategic leadership role. But I think it goes back to, I love this stat, only 50% report success. Well, what is success? And that is the big shift that I think has happened with CIOs that I think needs to happen with the chief data officers.
I can't take credit for this, it was from a McKinsey article in 2020 on reimagining the role of tech leadership. One of the big roles is translator. We have business initiatives, like business needs, like we need to have good loyalty programs or drive self service away from our call centres. Like, okay, well, how does technology achieve that? The CIOs had to become very business savvy communicators and learn how to communicate the technical work as driving the business results.
And I think from the chief data officers, that's something that they're gonna have to learn. How do I communicate the value of all of this data? What are these various roles, but more important than what are the roles like, what value do they add to the organisation? And so if you could say, for example, we have some data program and has improved our data quality and therefore reduced waste that we send, you know, we send marketing emails out to people that don't open them and we get the same response or even higher response sending 1/3 as many emails for 1/3 of the cost. Those types of things are what the chief data officers need to be moving towards.
What should the role of the CDO entail?
Vincent Yates (17:08):
Yeah. I think that that makes sense and resonates with what I heard. So I talked to a lot of people I was out there in Boston and I summarised a few technical notes, if you will, of what I think is the job description. And the one that you brought up just now, which actually came from our friend, Craig at McDonald's who is now the chief data officer at Best Buy - he said that chief data scientists are evangelists. Part of the job of a chief data scientist is, or chief state officer in general, is to be an evangelist. That is you have to inspire and teach your peers what's possible with this data. It's still a nascent technology. It still has yet to revolutionise a huge chunk of the business.
We've certainly seen it in marketing to some degree with the optimisations you've just talked about. We've seen it in some products with recommendation engines, but that's just the tip of the iceberg. And so really teaching others is part of that job.
Another one that I heard was about being a coach - that is, you have to not only inspire them, but then point out which ideas should we actually go after and how to think about that problem. Because what you're really doing in a chief data officer role is getting a business to prioritise data things over other business objectives. And I think there's a tension there, like who's gonna get credit? And so again, need to be a scrappy technologist. This came from the chief data officer Chewy, actually. And it was about how do you make the best of the infrastructure that wasn't really built for you?
And this is something we've talked about before with MLOps in particular. Chief data officers have big, the CEOs have big expectations of them in terms of delivering real value. And then they don't have the infrastructure and tooling that was actually built for them. And so that's tricky.
And the last one was around being a politician. And what I'm talking about here is really navigating those somewhat treasured waters of misaligned incentives with, by the way, again, most chief officers are new. Most of them are from outside the organisation. Most of them haven't been in a role like this before. Some do come from CIOs and CTOs, but most are outside. So you're expecting somebody who's new to the C-suite, new to this political water, to be successful in effectively taking credit for sharing credit for and getting budget from other executives who have been at it for a long time.
I mean, internally, it's me versus you sometimes Jason! Not really, but you know, I don't really want to go into that battle against you because you've been at this for a long time. You know, how to navigate these waters a lot better.
So those are the job descriptions in what I heard, but it made me wonder... Well, what is the responsibility of the chief data officer? What are the roles and responsibilities of effective chief data officers? And the first thing I want to start with is if you're responsible for the data within an organisation, what does that mean? Like what do you have to do? And one of the things that occurred to me as I was thinking about this is that back in the day when we were on-pre, all, all these things were on-prem, what did we call the buildings that held all our servers?
Jason Goth (19:59):
Vincent Yates (20:00):
Data centres, and who's responsible for a data centre, Jason?
Jason Goth (20:03):
That'd be the chief technology officer or the chief information officer.
Vincent Yates (20:08):
Exactly. Not the chief data officer. The chief data officer is not responsible for data (in fairness, we've moved to the cloud), but this is the tension. And so I'm curious how should existing CIOs and CTOs and up and coming CDOs be thinking about it, at least from your perspective? Who does what?
Sharing leadership with the CIO: Who does what?
Jason Goth (20:27):
That's a very good question. I mean, part of it is I think this was the topic of our last, probably very rambling, very unintelligible podcast, where I was throwing out some ideas that like applications need to be more architected around data and some of the data patterns. I think that is something that chief data officers can bring, like "hey, here's the more modern ways to collect processes, govern, and ensure the quality of data for all of these applications". And maybe that leaks into some of the application architecture itself. I think with all of these things, it's not going to be, you know, my favourite phrase of "Let's Jack up the radiator cap and slide a new car underneath". We're not gonna come in and say like, here's the chief data officer's role, just like we didn't do that with the CIOs... That transition happened over the last 20 years.
I think they're going to have to work through and figure that out and find out where the boundaries are between the two. But I do think data ownership, governance, and quality are big areas where the CDO can really help CIOs CTOs. And I think looking at some of these new data patterns and then working, we've talked about AI and ML ops a lot, like starting to work with them to implement some of those things. I think the CIOs can be a big benefit here, but we're gonna have to work to help chief data officers communicate why should we do this? And what is the benefit to that business? Because they've been, for the last 20 years, learning how to do that.
Vincent Yates (21:58):
Yeah. And by the way, when I was researching for this podcast, one of the things I saw under CIO as a definition was actually chief infrastructure officer. Have you heard that one before?
Jason Goth (22:07):
Never heard that one.
Vincent Yates (22:09):
Well, I thought it was interesting because I also came across a quote from somebody else who likes to break up CIO and CDO in the following. They say they like to use the analogy of a bucket and water. And in particular, the chief information officer is responsible for the bucket - that is they're responsible for making sure that the bucket is the right size, that there are no holes in it, that it's safe, that it's in the right place. And the chief data officer is responsible for the fluid that goes into the bucket, comes out of the bucket, that it goes in the right place, that it's of the right quality and is the right fluid to start with. And neither the bucket nor the water will work without the other. Of course, I think that's a very clean analogy, but I don't actually think I agree. I'm curious what your thoughts are.
Jason Goth (22:50):
Yeah. I don't agree with it either. Like that would tend to mean, in very practical terms, I design the data clusters or database clusters and you would design the data that goes in it. Right? I think that is actually a very poor illustration.
And that is actually the historical way we have broken down those two roles, between, let's say, technology architects and data architects, as we've built systems. And that's led to systems that are somewhat not optimal. So I would say a much better way to think about it is for some business objective, take any business objective, to reduce churn and think "what are the things we need to do to reduce churn?" Some of them will be very data oriented and some of them will be very system application interface and design oriented. So I could see the analogy working in that those two things, will have to work together. I think it may be more of a squishy boundary than people realise.
Vincent Yates (24:03):
Exactly, and I think part of the problem here is that as a chief data officer, I can tell you, I don't actually control the data that applications decide to generate. I can do my darnedest and I absolutely do. And we can write and prescribe guidelines and methodologies and processes and governance panels and MDM and all these things that we've been trying for the past decade. And it won't work. That's not how the world works. And in a world where there's more data than ever, everybody wants more access than ever, as we continue to push down this frontier, it's just unrealistic that we could have human-centric processes that control that. I mean, listen to our podcast or read our blogs on modern data governance and see all the challenges that you're going to face by doing that. So the idea that you could decouple a bucket, the infrastructure, from the data itself seems woefully inadequate. A bit naive, frankly.
Jason Goth (24:54):
My analogy for that is how many contact lists do you have on your phone? I've got the one I transferred over from my iPhone to my new Android. I've got the stuff I've entered in Gmail. I've got the one from my corporate email server. They're all wrong. I don't take the time to go through and make sure that they all stay in perfect sync with the phone numbers and emails, but eventually something aggregates them. In this case, my Android phone makes them available. And it's good enough. And so my analogy there, it's probably a bad one, but for these systems, yes, you're not going to be able... I've seen that many times where chief data officers have come in and laid down some mandates on all of the application owners and those mandates were promptly ignored because they have their own mandates from their people that actually do their reviews and decide their bonuses. But I do think they can work together to aggregate that data, clean it up, incrementally get improvements added back in, working with their technology friends and clean that up. I don't mean it to sound like it's a secondary concern, I think data is very important, but I do think it's going to have to be a gradual process.
Responsibilities of the Chief Data Officer
Vincent Yates (26:14):
Yeah. So if that's not the analogy, then I suspect our listeners will say, well, what, what is?
I don't think I have a very clear analogy, just to be really clear. So let me just try and articulate some of the things that I think this role needs to happen.
One is there needs to be some responsibility and accountability for the infrastructure. No doubt about it. I mean, look, if you don't understand what database your data's gonna flow in. If you don't understand what's streaming or batched, you don't understand the implications of that, then you can't solve the business problem at hand. And back to the point, part of what we've seen, both in the surveys that I've mentioned and what we've seen from our own client success and my own personal success is that you have to get crisp with the business early on and say, what do you as a business leader need for the data to do?
And part of that then comes back to helping educate them about what's actually possible. So the number one priority then becomes internal evangelism. And what I mean by that is you have to help them understand what's actually possible with data. You have to inspire them. You have to give them a path to go execute on that. And it's not just evangelism for the sake of evangelism. They actually have to action on this. Otherwise, you've failed at the evangelism. So the outcome of that, the metric around that is can you inspire a couple of business leaders to sign up for whatever you're pitching?
And the second part of that is evangelism externally. So we know that these people, these data practitioners are in incredible demand. US Bureau of Labor stats says data scientists in particular, which is only one of the groups you're going to need is going to increase by about 30%. The number of demand, that is, the number of jobs needed, will grow about 30% over the next two years. CompTIA did a survey and they have a prediction that it's going be 268 growth for data science of the next decade. To put that in perspective, cyber security, they think is about 253%. So data science is actually bigger than cyber, and we all know how important cyber is these days and that developers in general will be grow about 200%. So for every two developers you need over the next decade, you'll have to hire three data scientists. And again, that's just one of the neat people needed.
Last one I'll give you here around the top 10 tech jobs in 2022. Again, these things are all very subjective, but it just illustrates the point; one more way to triangulate here. Full stack engineers are number two at the moment. Data scientist is number three. Data engineer is number six. Oh, I forgot one, machine learning engineer, which is number four. So three out of the top 10 in-demand jobs go exactly to the people we're talking about, and I'd argue that full second engineering actually is a fairly data-intensive job these days, too. So maybe four out of the 10 are actually this.
That leads me to a place of saying that in ranked order, if you will, in front of evangelism, you have to find one or two business leaders and make them successful. And by the way, don't boil the ocean - only find one or two in the early days and make them successful. But that also requires you to talk about it externally, so you can recruit the people you need to actually execute on this. Which is really tough.
And then I think the other part of this is really the question of shouldn't include analytics or not. And in my personal belief is that it should. And the reason I say that it should is because at the end of the day, the value that chief data officers or data in general creates is by allowing the business to grow and potentially saving costs too, but really allowing the business to grow. And that requires analytics at the end of the day. I mean data by itself, we know how to do operational data stories. That's not new. I think a typical engineer would be able to pull up any random data field from anywhere in the company and show it on a dashboard. That's not really the big thing that's changed here. It's pulling lots of data sources together, doing something intelligent with that, and then showing it back with a prioritisation engine, etc,
Jason Goth (29:53):
It's funny you mentioned the security folks because it's not something I've thought about until just now, but I think CISO is probably a very similar role that is maybe a few years ahead of CDO in that life cycle. I think there's probably a lot of analog there too. For example, CISOs have become a very matrixed org with many different applications. There's some security policy, security governance, etc., that happens outside individual technology teams, but they tend to have representatives within those technology teams to make sure that those things are enforced. So they have somewhat of a matrixed org or a matrix function, within the technology organisation. Sometimes outside of the technology organisation as well. I think that is a role you could look at as a model - that is, "how do you work with the CIO and more technology driven functions of the company?" That might be a good one to look at.
Vincent Yates (31:01):
That's a really good call out. By the way, we have a CISO and I don't want to embarrass him, but my joke is that CISOs are really just a scapegoat... They're just there so that when you have a data breach, you can fire somebody, you can point a finger and fire somebody. It's not true, I realise, but the thing I want to tease out that though for real is a lot of how I imagine a CISO anyway, and you tell me if I'm right or wrong, is that it's really about mitigation of risk and that feels a lot like the cost cutting thing that we talked about before. It's sort of mitigating the damage of all this data stuff that you have to do or mitigating the damage of the IT stuff that you have to do. And the piece that I want to tease out is that from my perspective, I think a big part of what made CIOs and CTO successful ultimately, and a big part of what will make CTOs effective, is driving actual revenue. And I'm just curious, do you see the parallel in CISO or is that really just about mitigating risk?
Jason Goth (32:06):
No, I think CISO's job is to mitigate risk and, you know, from your description, if CIO means "career is over", maybe CISO is "career is soon over."
Vincent Yates (32:18):
Let's trademark that!
Jason Goth (32:21):
Yeah, sure thing. But I do think the analogy was one more of how to work with the technology leadership, not so much what you do. So I do think that the data focus needs to be maybe some, maybe the CISO is more about mitigating the rest of data security, data, loss prevention, those type of things. But I do see that the chief data officer more around how do we use data to improve company operations, improve services, improve service, improve the products, or add new products, those type of things. So it's very revenue customer focused, but I meant more as a way of starting by incrementally finding value, working with the CIOs as a way to kind of figure out where those boundaries of and roles and responsibilities exist.
Vincent Yates (33:21):
I think that's fair in that. I think that's a great parallel that we should probably investigate more. The other thing that came to mind as you were saying that though is part of what could happen here... For those organisations that are very far down the path, they've had a CDO for a long time. A lot of that was really around mitigating risk, that is governance compliance things. I wonder if it makes more sense to, again, remember that CDO didn't exist five years ago, for most organisations. Imagine that most of those organisations already had data. They already had applications. They had infrastructure doing all this stuff.
So this role really is some combination of creating a new role, but also taking responsibilities from other people. Part of that might be actually pushing that governance stuff that they've historically done over to the CISO, because it feels very natural - the compliance, regulatory, security... That all feels like it's sort of the same vein and era. What are your thoughts?
Jason Goth (34:11):
Yeah. I haven't really thought through it because again, I just thought of that analogy five minutes ago, but I do think that sounds very reasonable.
Vincent Yates (34:22):
Off the cuff. You don't want to just agree with me. That's too dangerous. Clearly, you've worked with me before!
Jason Goth (34:27):
But I do think one of the things we often talk about and we've written about is with data, you want to start with some real problem. Don't just come in like, "Hey, here's the new org and here's what I'm responsible for!" Let's start with some real problems. If your organisation's biggest problem is data quality, i.e., we don't have good data and we can't report on anything, well then let's start there. If your company's problem is we don't have enough data to make decisions, then let's start there and let's evolve our way to what the right relationship is with others.
Vincent Yates (35:09):
I think that's good. I'm getting a look from our producer, so I have to be a little more punchy here in my last little bit. But anyway, that leads me to the question, as you think about the CIO/CTO role, do you have a perspective on if it's better to have two roles or one role? Back to my question, should it include analytics or not?
The importance of analytics
Jason Goth (35:27):
I do think it should.
Vincent Yates (35:28):
Okay. So you agree with that. And should I point to CIO/CTO? Have you guys resolved this or, or no?
Jason Goth (35:38):
Yeah. CIO/CTO, I don't know that there's a right answer. Whatever companies do, there are these internal and external roles and whatever they call them, those roles are fairly well defined. Whether that becomes two different roles or one, who knows? I think it may be all different in different companies. To me, it's more about, do we know what they're doing and how they interact with everyone else, specifically the technology organisation.
Vincent Yates (36:08):
Okay. That's fair. As you guys (CIOs) moved from cost saving to real revenue growth, how did you manage the incentives? What I mean by that is if you go a business today, let's just make it easy and say we go to marketing and we say, "hey, we can help you reach three times as many people or have three times the click through rate, which means you can decrease your cost by a third and keep everything the same." They say, "that sounds pretty good. What do I have to do?" And you say, "well, instead of hiring that new marketer you wanted, or instead of spending all that on social or whatever you're doing, spend it on our people instead. Give me that money and I'll get all that for you." That executive may or may not be excited about that proposition. Every organisation deals with the internal transfer of money and politics and KPIs a little differently. Are there any good lessons to be learned from your experience here?
Jason Goth (37:03):
Well, my experience with most of my clients is those mandates generally come down from above, right? Like, "hey, CIO, figure out a way to drive three times as much revenue through our digital channel." And so I don't think that the business really cares that much about how we go do it, they just want it done. So to me, those tend to flow down. I think that the question will become for the chief data officers, are there things like that that would flow directly to them or would they be a part of a lot of other initiatives? I think at least initially, they're gonna have to be part of other initiatives because they're going to have built up their infrastructure. And infrastructure, not in terms of like servers and data, but in terms of ways of working and social capital within the organisation, that type thing.
To where people will then come to them and say, "Hey, Vince, figure out a way to X." So I think it's an evolution. If you think of it as a path up the mountain, I think CIOs are closer to the top. CISOs are somewhere in the middle, chief data officers down to bottom, but I think it's a similar path and, and people will start to engage just like they start to engage CISOs directly now. I don't know that we should try to design what those things are. We should let it evolve into what it should be.
Vincent Yates (38:30):
No, that's totally reasonable. Sharing a few statistics to back up what you just said. Among organisations that typically have a CDO, we're looking at roughly 40% that feel that they have a CDO CDAO role that's established and about 43% say it's nascent and evolving. So that fits exactly what you said. And again, 17%, they're struggling with the turnover of this role. Not terribly surprising.
Jason Goth (38:54):
See, I give opinions and he gives data. That's kind of the way we work together.
Vincent Yates (38:57):
That's the difference in a CTO and a CIO, that's awesome.
Okay. So with that in mind, in this incentive thing, the next question I think follows naturally (although I've maybe biased the listeners a little bit just by the ordering here), but to whom should this person report to?
Should the CDO report to the CEO?
Jason Goth (39:19):
Pass. I think that's gonna be my answer.
I generally don't know. I think there's probably better people than me to answer that question. I tend to not focus too much on who reports to whom.
Vincent Yates (39:33):
Well, but it matters, right? I think you hinted at it before with the CTO and where they typically report. Let me make some points here. We know that fewer than half the organisations that are already sort of down this path actually have a data-driven strategy. If you want your company to be data-driven, you want to extract the value out of your data, then you have to be a data-driven company. In order to have a data-driven company, that means your corporate strategy must be heavily influenced by what? Data, right? Therefore heavily influenced by whom? The CDO. We just said that the role of a CDO is to effectively influence the culture. Secondly, if you think about that remit of how do you transform the organisation, your point is right. It comes down from the top.
And so to me, it seems that the CEO and the board have to empower this person to actually go carve off the pieces that are relevant. Because again, the organisations that exist today, they have data today, maybe not using it well, but they have it today. The businesses operate today. Maybe not as effectively as they could, but they do operate, which means that this new role has to carve out functions and authority to do that from a variety of different parts of the organisation.
Jason Goth (40:51):
Yeah, I agree. I think my reluctance to answer is more out of a, well, one, I don't want a bunch of chief data officers coming and finding out where I live, but it's more out of a practical aspect of I don't know any CEO that wants more direct reports. And they're going to only accept one very begrudgingly when they realise, wow, this person adds a lot of value and needs to have a seat at the table. My answer is more like, that is where definitely where you end up...
Vincent Yates (41:35):
But maybe not where you start.
Jason Goth (41:35):
But not where they start. Right.
Vincent Yates (41:37):
Okay. Well, let me try one last pass to convince you real quick. We said earlier that CDOs are ultimately responsible for the most valuable asset in the world, that is the data. They're expected to leverage it to bring about major transformational change. They're given remarkably little time, we're talking 18-24 month kind of timeframes to do that. And despite that audacious expectation, they're expected to do it without reporting the CEO, according to you? I just don't know how you have the authority to restructure orgs, make technology decisions, how to change incentives that get the budget to re-platform in the cases that you need to, let alone force somebody to implement data changes to capture the right data.
Jason Goth (42:20):
I agree. You know, those are all things you need to do. I just think the way to do that is not to set someone to immediately be in a knife fight with the CIO, because that is actually going to make it take longer. It's like, hey, if this person can work with the CIO... To me, it's the same as the architecture team. And I know some companies who have a chief enterprise architect, which doesn't report to the CEO. Same way with chief digital officer, chief marketing officer, like those roles end up generally in a fight in the parking lot. So I think there's a lot of of practicality in that argument.
I think a better thing would be smart CIOs need to learn to listen to the data officers and carve out those things you just said, instead of viewing CDOs as a, "oh, here's my guide with all the Excel spreadsheets." Right. You know, start listening to them as they would some of the architects. But I do think over time, the more and more footprint you have within the organisation, the more and more, it may be something that you can do, you know, directly.
Leadership through influence or authority
Vincent Yates (43:36):
Yeah, that's right. Last thing I'll say... I think you're right. There is this question of, do you do it through influence or do it through authority? And maybe you're right. The early days maybe influence is going to lead to a better outcome versus cramming it down everybody's throat initially. That could be right.
Last thing I will say though, to your point of CEOs reluctantly taking this on. In reporting to the street, so financial reporting, you're seeing an increasing amount of data being pulled into investor reports and an increasing amount talked about on investor calls. So this one comes from PWC, they did an analysis of the top 2000 company's reports over the past since 2017. And they analysed the number of terms, analytics, machine learning, data, AI and a typical company now refers to data 52 times in their annual report.
And while the range fluctuates dramatically between companies from just a couple mentions to as many as 200 times in a report, the trend is that 2/3 of all companies have grown in the amount of times they mention data or data related assets since 2017. And so my only last point there to CEOs, boards, and investment teams is that the street really cares about this and they recognise this will be a differentiator in the long term and they're expecting CEOs to address it. And so having somebody who understands that and is mind melt with the CEO, at least from the strategic standpoint, I think is going to be important. Now that they have to report to them? Maybe not, but they are going to have to increase the amount of time they spend with these people.
Jason Goth (45:16):
Yeah. I agree with that.
Vincent Yates (45:17):
Okay. Last question then really is, imagine that we're talking now to the roughly 80% of organisations that don't have one of these They're thinking about, "hey, maybe we should do this." I'm curious, should it be an internal or external hire? How did that play out for CIOs/CTOs?
Sourcing a New CDO: Internal or external hire?
Jason Goth (45:38):
I think it depends on the company. It's more of a talent issue. Do you have the talent or do you feel you have the talent? If you do, internal is great because listen, there's a lot of institutional knowledge in the data folks heads, you don't want to lose that. But you don't have the talent, then you definitely want to go external and get it because it is such a critical thing. I would argue that probably many do not have that capability internally and should go external. But it's hard to say,
Vincent Yates (46:13):
I think that's reasonable. If you look at the data again, roughly 50% hire an external change agent. I mean, you realise that you are asking for a transformational change. Sometimes it's easier to bring somebody new in and other times it's actually worse. This body rejects this foreign body and makes it a little bit harder. And by the way, the research from Randy and New Vantage, 91% of CDOs believe that cultural impediments are the biggest blocker. It used to be in like 2017 or something that it was typically technology. I think the cloud has just moved along far enough down this path that it's no longer technology, it's actually culture and people and process. That is the biggest blocker. And so having somebody who's savvy becomes really important.
Jason Goth (47:00):
And that's a really, you know, becoming more data-driven is, is a really big shift, a much bigger shift than people believe. Say we're going to go do this programme, whatever programme it may be. We're gonna go build a loyalty program. Well, that's going to take like a year. Can you break the problem down a little smaller and do it in a lot of incremental steps that you measure and data drive? The idea of build, measure, learn... The lean idea, build, measure, learn. That requires the entire company to shift culturally, not just to be more data-driven, but to do software delivery and other things in a way that you can be data-driven. And so that's where I think the CIO partnership there, like whether it's a reporting relationship or not, there has to be tight alignment, in that we need to work together to move to these new ways of working that are very iterative and data-driven, as opposed to we do a study and the highest paid person in the room announces this is the new three year strategy. And it's three years from now before we know if it works or not.
Start small, find your people, share your wins
Vincent Yates (48:14):
Right. I think that's exactly right. And just to quickly summarise, I think a lot of it is exactly what you said. You need to come in, regardless of internal or external, and realise that there is a big shift that has to take place. You have to ultimately start small, find a few people that are "green dots." That is if you have an org chart and you show who's aligned and who's not... The green dots are the people that you are more likely to work with you. Start with some green dots, make them successful, start small, and don't boil the ocean. Don't try and think through everything that we might do someday. Just think through the stuff that we have to get done to make that one person successful and then show the success and leverage that to sort of continue to grow it.
And that really means focusing on the business outcome, staying hyper focused on how to actually drive value. Talk about that, focus on that, drive towards that. And evangelise that once you actually have it.
The last bit would be to remember that you're not alone. I think the most encouraging thing to me as somebody who's had this role at many companies and felt very isolated and very alone, very much like I was trying to create the wheel for the first time on my own. Remember that this is normal. This is common. This role is new. These people haven't done it before. The success is still very early days. It's not easy to do this well. And so create a community around you and find other people like you.
And if you want do that, you can always reach out to Credera. We have such a community that you're welcome to join. And this event I went to last week at MIT, this CDIOQ seems like a really great community. And I'm sure there are many, many more just to be clear. But do find that community. Any closing thoughts from you, Jason?
Jason Goth (49:58):
No, it's a good topic. I enjoyed this one.
Vincent Yates (50:00):
That's good. Well, for those of you, who'd like to learn more please visit the insights page on credera.co.uk. Thanks for listening. And I hope you'll join us again. Thank you, Jason.