The Wrong Mandate: Why Boards Set Data & AI to Fail with Kyle Winterbottom

[Music] Hi everyone, welcome. It's time for Catalog and Cocktails. You're honest, no BS, non-salesy conversation about enterprise data management. I'm Tim Gasper, joined by Juan Sakeda. Hey Juan. Hey Tim, it is awesome to be here live. It's not Wednesday. It's Thursday. It's not our usual time. It is actually cocktail time in the UK. But I am super super super excited to have Kyle Witterbottom who's a CEO and the founder of Orvician Group. If you're in the data space and you are a data leader and you don't know Kyle, I'm like, "What the heck, where do you live?" I think Kyle's post show up on my LinkedIn feed every day and I'm always thinking and he always, I think you live this whole honest no BS herself all the time and I love what I, I look at your post, I look at the comments and sometimes like, "Oh, I got to get some popcorn because this is going to be interesting." Anyways, Kyle, it is super freaking awesome to have you here on the podcast. Finally, how are you doing? I'm very well, thank you. Pleasure to be here. When you asked me to come on one, I was like, well, if there's a podcast or the name of a podcast that I guess summarizes my approach to things, it's a non-salesy no BS podcast. So yeah, there you go. Do I have to be here? Alright, we'll do it, but before we jump in, it is early in the morning live here, but I'm curious, what are you drinking? I don't know, it's 4pm your time, but what is your cocktail beverage of choice in nowadays? Coptale beverage of choice. I'm a whiskey drinker, so if I have to go for a cocktail, it'll be an old fashioned, but I'm a big Belgian beer drinker if that permits, not 4pm on a Thursday unfortunately, although it's right before the Easter weekend, so we're probably not too far away from that point in time. Eddie, what's your favourite Belgian beer? There's a beer called Triple Carmelite, which I run into when I was out in Brussels and had far too many of those considering they're 8.8% brief, but yeah, that's another story for another day. I've actually tried that beer before, that's a really good one. I think Delirium Tremens is probably one of the best beers in the world for me, but I think also one that really hit home. I think one of the first Belgian beers I've ever had that have interesting memories giving me what a drink I made of those, so the Golden Drock. That one, so yeah. Anyways, we got a lot to chat about here because when we posted about this, people already started to go chat. Honest, no BS, what do you mean that the people in the boardroom continue to design the wrong CDO mandate? Yeah. Well, let's start as we mean to go on, shall we? So I think it's pretty obvious right now, and I know there's a lot of talk about this topic, but fundamentally, however long organizations have been doing data, data management, analytics, now AI and all that kind of good stuff, the talk is still, we're not getting value from it that seems to be a very small finite number of businesses that can point to demonstrable, tangible commercial value that they're getting from a lot of these initiatives. And when you trace it all back, in my opinion, most things are downstream, most of the issues in our industry are downstream from the fact that the people in the boardroom build the mandate of the CDO role or the de facto CDO role, whoever it is, the most senior data person, whoever is accountable responsible for the data capability and function inside of business, nine times out of ten, it is not aligned to delivering value and outcomes. It's aligned to delivering frameworks and models and dashboards and reports and platforms and lakes and lake houses and swamps and blah, blah, blah, blah, blah, blah, blah. And then they get really annoyed when that's what they get. So we're in this space where the mandate is, you know, it's never aligned to delivering the thing that they fundamentally want, even though they don't ask for it and end up firing people 12 of 18, 24 or 36 months down the line for not delivering a commercial ROI. Now, I think that the point to note is that the actual, the mandate itself is normally the symptom of a bigger problem. And I think, my opinion is that the root cause is the perception of what data is or it exists to do for an organization in the boardroom. And nine times out of ten, you know, I'm fortunate to speak with people, you know, the board members every single week, because obviously what we do as a business and fundamentally, they see it usually as a technical problem or challenge that needs fixing, addressing, you know, as you'd expect. Data needs to be moved, it needs engineering, we need to visualise it, needs dashboards, you know, blah, blah, blah, blah, tools, technology, infrastructure. And that perception leads them to create the mandate, which they create, which leads them to not get any value out of the other side of it. Now, that's not to say that all that stuff is important. It absolutely is, but that's kind of table stakes, in my opinion. This is a very direct and honest statement, I think, about the state of affairs right now around the role of the CDN. While a lot of them fail, right, it's not a high success role and it's a very high turnover role, right? You know, is this categorically a problem with like the fact that this even, this is perhaps a very edgy question here, the fact that this role even exists, like, you know, because I think there's a lot of feeling right or wrong that like, well, you know, the CRO is in charge of sales and the chief customer officer is in charge of the customer experience. And so the chief data officer will be there in charge of the data, right? And so there's like kind of a boxing in that happens. You know, is this like a categorical issue or do you think we can actually change the mandate and find a better way to direct CDOs? I mean, you can definitely change the mandate, right? Because a big part of what we do as a business is try to start at that point, you know, often businesses will come to us. We're looking for a CDO here's what we need and there's a lot of unraveling on our side as to maybe don't need that like what what your goals and objectives, what the KPIs that you will use to manage those goals and objectives, what the decisions that you will make around those KPIs that dictate the goals and objectives and so on and so forth. So there's an unraveling of that. So the mandate can definitely 100% be changed because we've done it with good success. And also there are businesses out there and some that have been very successful that are adding eight, nine figures annually to their top and bottom line through using data to drive better decisions. So not many of them, but there are some. So it proves that it can be done. But is it a all-encompassing problem in the for the majority of businesses? I think it is. You know, if you were to look at the amount of businesses doing something with data, there's not many businesses out there now that the aren't obviously at varieties of different size and scales and complexities of course, but the number that are getting commercial value out of doing it is very slim in comparison to the number of businesses doing it. Tim, you mentioned something then that I think is really interesting and a good point to touch on. And I've been a big kind of believer in this for a long time. I think a lot of businesses went on this journey to being data driven in quotation marks because it was the right thing to do. I don't think that anyone doesn't believe in it. I think they believe that there is value there. And in all honesty, that's how most of the conversations I have start, right, especially businesses that have never hired a CDO is usually, we've got all of this data. There must be some value in here somewhere, but that's the Stein point. So they go on higher somebody to try and get them to figure out where is the value and part of that conversation comes. Well, we're trying to figure that out. We're going to need a platform. We're going to need this. So let's start building. I think a lot of the time businesses get sucked into this. And we're seeing the same thing with AI now, right, even though AI is not new, the whole Gen AI and other M pieces, relatively new and it's become more, I guess, commercialized from that lens. I think businesses got sucked into this because there was pressure to be perceived, to be seen, to be doing something. You know, almost that the cost of being seen not to be doing it was more detrimental to the brand and the perception in the external marketplace that it was wasting the load of money doing it wrong with. I think this, a little bit of my, the historical soapbox I get to is like, I honestly think that the Hadoop era, the big data was kind of like a decade, a lost decade there because, hey, people didn't actually need to use Hadoop at all that stuff. But they did it because the CIO was in a meeting and a dinner and they're like, oh, we got her, we're doing this stuff. I'm actually that. I have a dinner again in a couple weeks. I better say that we are doing that stuff. And then I think we've kind of gave that perception, but because we've been doing data before, like, I mean, there's data warehouses and you're doing analysis and all this up and you are finding trends and like, we've always been doing that type of work. But I feel that what has been, We have this focus on the the technology on this on data. And I actually, so like I push this out, I'm curious what you think is companies, they have a data problem, but they don't really have a data problem. Like they have an app, but they don't really have an AI. Like that's not really a thing yet. You have to go focus on as an organization. 'Cause those are means to an end, but you have is a work problem. Like what you need to go do is figure out how to go do more of your work efficiently, do more with less full-time stuff. And obviously data is gonna help me AI is gonna help me. My context is gonna help that. My insights are like, so I think a lot of the framing I've been kind of talking to data leaders. And I'm having good positive perception. I'm curious because you speak also to a lot of them is, "Look, you have a work problem, not a data problem, and you need to get out of your bubble." Because at the end, if you're going back to like, if you're the CEO, you own data. I was like, "Yeah, you own data, but data by itself isn't the goal. You need to be able to take that and tie that back to your different business line. So you are completely cross-functional with these things." Now the issue is that being cross-functional, like that means that you are stepping on other people's toes and stuff. And that's, you need more business acumen and be growing the politics of stuff, which a lot of the data leaders are more from the technical side that they would do that. Anyways, I'm ranting here, but does this resonate with you? Does it resonate with the leaders that you're seeing in the stuff? Curious. - Yeah, yeah, 100%. I think there's a big portion of everything that you've just said that is absolutely true. I think a big problem here is accountability, in terms of if there's a disconnect. So often what I see happen is people in the boardroom will appoint a CEO. They view that as an enabling or supporting function. This may not go down too well, but I think one of the worst things our industry ever did was kind of subside to the fact that we are an enabler. And there's a lot of CDOs that will turn around and say, "Actually, as soon as you're seen as an enabler or other support function, you kind of, you're done." Right? So your first point, a business doesn't exist to do finance, unless it's obviously a financial service in the shoes, it doesn't exist to do HR, it doesn't exist to do data. It exists to do whatever that business does and do it better and do more of it and blah, blah, blah. So if you sell clothes, the way you improve your business is by probably selling more clothes, different products and blah, blah, blah, blah, blah. Right? So every function in a business should serve that purpose. Right? Data is no different. The data function should be how can this capability that we are building enable this business to sell more of what it sells, do more of what it does and improve the performance of the overall business? But there is an accountability problem, in my opinion, in that board's higher or see here to come in, they are hired to do data and all of the kind of technical technology infrastructure type things that support that to enable. Right? And at that point, often boards think that just because they've done this stuff, that the value is just gonna appear. Like someone's gonna go and turn on a light switch and dollar bills are gonna start falling from the sky or something like that, right? And we all know that isn't the case because something needs to happen, just because you've got a warehouse or a lake house or a platform doesn't mean anything. That work is important and it's valuable, of course, but it doesn't drive value until something happens with it. And I think that's the big problem because there's a disconnect then, because equally, your sales leaders or whoever, they're also not accountable for making better decisions with data. So if the data leader's not accountable for driving outcomes and the business unit leaders aren't accountable for driving better outcomes by using data than somewhere in the middle it's getting lost. - I think just to add to this a little bit, I think one thing that's challenging is, you know, what are those KPIs that a chief data officer should impact, right? Because I think let's compare again to like a chief sales officer, a chief revenue officer, right? Like, of course they would say like, oh, of course, like I'm not measured on my ability to implement a CRM, right? I am measured on my ability to drive sales, right? With a chief data officer, I think they would kind of agree that like, well, yeah, of course my value is not to bring in a catalog or bring in snowflake or data bricks or something like that, right? My value is to move the needle on dot, dot, dot, right? And that's where I think going back to what you said about data-driven culture, one thing that I think CDOs found a bit of an opportunity around whether it's valuable or not, right? Is like, well, how many people took the data enablement course and how many people are, you know, all even point of catalogs a little bit and how many people are logging into the catalog and they found these metrics that they thought maybe are kind of related to enablement or something like that. But are they really valuable versus, you know, what should the KPIs be? I mean, look, that changes by context of business, right? But those KPIs in my opinion should always be related back to what are the goals and objectives of the business? And fundamentally that will come down to normally increasing revenue, profit, sales, reducing cost, improving operational efficiency, mitigating risk. - We're going back to your close example, right? Like, there should be a. - Sorry. - I'm just going to learn. It should, going back to that clothing example, it should go back to the accountability. They should actually be willing to sign up for what are the key metrics related to the efficiency of how you're producing clothing or the amount of sales around clothing, right? - You might have inventory, you have your amount of reduced cost by, you know, better forecasting on inventory so wasteage and all that type of stuff. There will be headline goals and objectives of a business and then the layer beneath that is the KPIs and the metrics that they use to determine whether they are succeeding or failing against those goals and objectives. And it's those KPIs that CDOs need to get to. The problem often is, is they're too disconnected and too far removed from that, you know, because of the mandate, because the mandate has dictated to them that your job is to come and build a platform and a governance framework and, you know, devise a data strategy and an operating model and they do that and then they get fired because they've not, you know, they've not delivered any ROI and it's like, well, no, wonder they've not delivered any ROI because you didn't hire them to deliver ROI. And you know, businesses only get concerned about that after two or three years in roll and they've done, you know, several tens of millions of investment into people, tools, salaries, new technologies, new platforms, all of this stuff and then look at it and go, what have we got out of this? And it's like, well, you were never going to get anything out of it because there's a disconnect between what you're asking these persons to do and the outcome that you want to drive fundamentally. So this is, so your position here is that this should be then what the board should be mandate, right? The higher level of should be saying, look, we're hiring you, you're hiring a CDO, not to go do all this technical stuff because we need to increase customer satisfaction because whatever is our top mandate of that. So how are you seeing this shift right now? Are you seeing, are you seeing boards actually thinking about this or is this shift starting to happen or are you still, man, screaming in the cloud, like, come on people, I say this a lot, we're having this conversation which sounds freaking obvious. And why do we have to state the obvious over and over again? - I think, you know, I think that comes back to the accountability and the disconnection piece, right? And you know, the wrong mandate is merely a symptom and the symptom is that the perception of what the role exists to do isn't aligned to outcomes. They don't think, you know, in most cases, in my opinion, most boards don't look at this. And I know there's a big argument that anyone in the boardroom will sit there and say, well, everything we do should align to an outcome. And that is in theory, very correct. However, in the data, little ecosystem bubble that we find ourselves in, that's often not the case, right? Because they perceive this role to be, it's a highly technical role. We need to fix our data quality. We need to find ways to move it around the business. We need a single source of truth. All of this stuff that is absolutely valuable, but on its own, it doesn't deliver value. And I think there's two, there are two different things, right? Work can be valuable, but it doesn't necessarily drive value until something else happens with it. So to answer your question one, how we see in this shift, we are, but it's slow. You know, it's turning slower than the Titanic would. It's probably the right way to frame it. There are businesses now in organizations where we're getting into conversations with where the mandate is shifting, and you can see that. And I put a couple of posts out on this recently where the way to spot this is when businesses start to look for very specific skill sets. So as an example, a business is looking for a CEO. Yes, they need the CEO to do all of the general end-to-end, you know, CEO stuff. They need them to build a team. They need them to devise a strategy. They need them to be able to build a platform and blah, blah, blah, everything that comes out of that. But you will also then start to see something like, I don't know, we want someone that's got a strong foothold of machine learning into pricing, pricing optimization, or something around that into, I don't know, maintenance improvement. So you start to get into quite a specific domain experience. And what that usually tells you is that they figured out the value of Now obviously every business as it relates to data can have multiple value levers and often does. But usually when you start to see that it's like they've figured out, hang on, this is the one area where data is making goes money and a lot of it. So how do we make that bit even better? So all of the other stuff is still important, they still need reporting, they still need dashboards, they still need insight, blah, blah, blah, blah, blah. But can we make this bit better because this is where we are winning, this is where we're taking market share of our competitors and blah, blah, blah. But those conversations are still fewer and further between than they should be. And as you said one, this seems really obvious, right? It's almost a bit of a head scratcher because you've got very smart people that sit in a boardroom that run multi-billion dollar organisations and yet this start you still turn out firm in a conversation about something that seems ridiculously obvious. Okay, so what I find really important about this trend that you're seeing, that you're describing is the focus on the domain. And I think this is also something I've been observing actually just in general, in this new AI world that we live in and I'm telling people like, look, a lot of that work that you're doing, that operational work that you're doing is getting automated, right? Now that means that you're going to be focusing on the on small amount of operational work, which means that you have the opportunity to become more strategic. But being strategic, like you can be a generalist or a T-shaped or a go wide or you're going to go down deep into something. I think this is the moment that people say, like, you should actually have to take the opportunity to go deep into some level of expertise. Yes, technical could be, but now we're talking about business wise. What is your domain? And this is starting, this is something I'm starting to see too, like, if you actually have data leaders, and that's just practitioners too, but it's like, oh, I know set of teams in the platforms, like snowflake for any industry is the same thing, data breaks, like the catalog is all the same thing. It's like, how it's being used to be able to go to something very specific for the industry. So that's the, that's I'm so, anyway, I'm just saying like I agree with you. And I think something I want to highlight more is that it's this industry aspect. So two things I want to bring up is one, somebody here actually good comment is why would a real CDO accept the role of a CDO where their real goals are not laid out? And it seems like that could be like, that's interesting question right there. But then also is this the push to saying, hey, data leaders, not only do you have to have a business act, you know, like understand how business works in general, but also you understand the industry. And then does that mean that we're not going to start kind of having specializations in in CDOs and stuff. But then I also wonder is like, it does this happen like going back to the sales and the CROs and stuff. Are you, if you go, you become a CRO a two sales officer, are you specialized in a particular domain? I think probably usually people's careers are and maybe, and then you see later on, they make a bigger shift in their careers. I'm going to take this and sometimes that works well. Sometimes that work, that doesn't work at all, right? So yeah, anyways, I'll shut up now. So, to answer the question of the person that asked the question, it's a very valid question. The answer is, well, you'd have to ask the people that keep accepting these jobs, right? First and foremost, the second part of that is it's not that the goals are not clearly laid out. It's just that the goals are normally, I refer to them as outputs versus outcomes. So they're outputs. Like the goals are we want you to do it to devise a strategy. We want you to build a platform. We want you to implement a governance framework. We want you to implement a catalog. So people come in, that's kind of par for the course, right? So people come in and go, I can do that. And they go and do it because they've delivered what they've been asked to deliver. And I've said this very publicly before, nine times out of ten as a CDO, in most organisations, if you deliver what you're asked to deliver, you will probably get fired. Because then what happens is they turn around and say, that's great. And it hears you bonus because, yes, in theory, you've delivered exactly what we've actually delivered, but we don't need you anymore because this is cost as a load of money and we've not seen any ROI from it. So it's back to that point where you would never going to see ROI from it, right? So it's not that the goals aren't laid out, it's just that the goals are never linked to business outcomes and value in that sense. The part around the domain experience, so this is something, you know, there's always it's cyclical, right? You get big swings in this centralized decentralized, Turban's Moe call this type of stuff. We are definitely seeing a shift towards more decentralized types of organisations now where, you know, and again, typically with more mature organisations where they are starting to hire like why would classify as like a mini CDO role. So remember like head of data that sits within a specific business domain or product line. So like each domain, how a product line has their own head of data which acts as a mini CDO, and it's pros and cons to all of these operating models as we all know, right? But that is definitely something that starts to starting to become a little bit more apparent. And I guess to your point one, like if I look at most resumes, like there's very few that talk about the domain or the product line that they operate in, you know, it's normally like you say it's quite holistic in terms of my skills are within this. I managed a team. I was responsible for a cloud migration project and blah, blah, blah, blah, blah. And it's very whittling there around, you know, and I did all of that work into revenue management or whatever the case may be. And I reckon we're probably going to start to head to a place soon where people will start to realise actually a lot of these roles are becoming a little bit more specific and specialised. And I'm going to have to put some eggs in a basket around where I'm going to apply my trade. I think that's what we'll start to see. Yeah, I think it's interesting, some of these shifts that are happening and in your right, there's always this pendulum that's swinging back and forth. On the decentralization thing, I've been seeing that a little bit as well where, you know, obviously you want like some of the more like platform oriented investments and, you know, kind of overall policies and things to be a little bit more crosswise. But connecting back to something you said a little earlier when you're like, "Hey, you'll start to see that people are finding the value lovers if they're asking for a head of data or a CDO that has experience in, you know, machine learning for pricing optimization or something like that because that's an indicator that they know what the lever is around value. Sometimes those levers can be department or business units specific, right? And that kind of ties in this idea of like, "Hey, if you have a head of data that's embedded in a business unit or in a particular function, then maybe they're actually going to be a little closer to the action and be able to take on accountability a little bit more directly than somebody who's a little bit more centralized a little bit more removed." Yep, yeah, 100%. And that's what we're seeing start to happen, you know, in that example, that was the exact situation where it was like, you know, they believe that they win, this organization believe that they win in that specific domain around pricing in the, in the kind of sector that they operate in. So they already think they're market leading, they know there's a chasing pack and they're like, "Okay, well, how do we get someone that's really come from that world?" So they are big enough and have broad enough experience that they can manage the entire data analytics organization, which is fairly sizable and they still want all of the other stuff that comes with it, but actually how can this person really drive that value lever forward to optimize it and make it even better than it already is to kind of, you know, keep water between them and the kind of the chasing pack from a kind of competitive advantage standpoint. You speak to so many boards and data leaders. What are the patterns that you're seeing on kind of the organizations? Like, so where is the reporting lines? How are the teams? If it's decentralized, how does it look? Centralized, how does it look? And what are the patterns that you're seeing that that's working? I know there's a lot of context, I think, depending on what works, what definitely doesn't work out. I'm just curious to kind of get your insights on this. Well, I think this is one of them questions that in theory, reporting lines shouldn't matter, right? But nine times out of ten, it usually deals with things. Oh, thank you. You flow straight back into what the perception of the role is. So there's been a bit of a, you know, and again, these things can be cyclical. There's been a bit of a pendulum swing, you know. So most, as we know, most data analytics functions were born out of IT, right? So they usually stand at the CIO. There was a spell probably four or five years ago where there was a bit of a shift out of out of IT because, you know, again, the complaint was still, we're not getting any value from this. So let's try to put them within the business. So a lot of CDOs then moved, you know, in some instances, very rare, but some instances to a CEO, some to a COO, some to a CFO. You even had some businesses try like a chief transformation officer or a chief innovation officer, something like that. So the reporting line went, went through that, you know, a two to three year cycle and then the complaint was still the same. You know, we're not getting any value from this and have shifted it back. I think a lot of the time because of, again, to go back to the mandate, when a business has got a CIO that's building infrastructure and technology and tools and all that kind of stuff, and they've got a CDO that's seemingly doing the same thing. People in the boardroom are looking at this going, well, why have we got two CXOs doing what to them appears to be the same job? And it's costing a lot of money and we're still not getting any value or ROI out of this. So, and, you know, let's be honest, most CIOs, well, in most businesses, the CIOs probably got more stock than the CDO has. So there's been a bit of a trend where businesses are getting rid of the CDO, given the strategic responsibility of the CDOs roll back to the CIO and then that's kind of creating this layer down level of a big hiring spike actually because obviously the CIO can't manage the day today so they go and hire a head-of or a director or a VP to manage that. So we're seeing a lot of that. So really in the majority of cases I think the CDO role or whatever the title is is still reporting to the CIO in most instances and obviously there's a lot of, in my opinion, a lot of challenges with that because that just usually kind of drills down even further into the fact that it then just becomes a technical technology related role. It's a very function, right? Because at the end if you really want the data, data teams and data leaders to be able to provide, we focused on the pricing strategy and the stuff and then they're kind of pushed down in the CIO kind of seems a little bit of a disconnect. So how much are you seeing then? I don't know if it's more in the decentralized, where you know you actually have core data people in a pricing function and finance or have multiple data teams and multiple heads of data around there. Are you seeing that then if so like in how are they coordinating and seeking things? Yeah, I mean context is massive in this, right? Because I think every business is different. So in their examples where the CIO still has the central kind of strategic accountability and responsibility and then they've got heads of data in different kind of business domains or kind of product line functions that are reporting to them. In some instances you have the same setup but there's a CDO there and the CDO still might report up to the CIO. So you've just got layers upon layers. It's one of them, in theory there's no right or wrong answer but often there is a right or wrong answer but I think the decentralized thing is starting to change how organisations are looking at this a little bit and I guess the more mature businesses now there's almost this matrix organisation where they have, you know, they're reporting to the business leader of the domain but with a dotted line back into the central capability whether that's the CDO or the CIO. So it is changing a little bit and I think that's only going to be a good thing but there will be many people that will argue and as with anything there's pros and cons, the cons of having decentralisation is that people go and kind of create their own fiefdoms and this mass duplication of work and so on and so forth, right? People will have an opinion on what's right or wrong but yeah. Yeah, you know because some of the patterns that I've seen here is, I think it's a little less what you're saying but explicitly is if you have a central data team, right? But then those are the ones who end up becoming the reporting function and stuff like that and people complain about like, I don't see the RRI, just generate a bunch of dashboards type of stuff and then there are the ones who are very core small central because they really have embedded kind of data places over it so then they end up being kind of that dot, they end up reporting to a business function but they're dotted line to central. Also that is because people thinking about career growth is like, wait, I'm a data engineer on analyst but we're going to go, I don't want to become a pricing person or something but so they want to have that dotted line. Then I also see kind of separation of all, we're the IT so we'll do the backend tech stuff but we'll do all the data engineering and all that but then somewhere else is like, no, we're more the customer facing side so we do the analytics of this but then then you got this gap between, wait, we're data teams but we're completely separated so I'm like, I get this data that I don't understand what it means and then that, oh, you get some, you get some word cost dynamics too right because in some organizations there's a lot of scrutiny on centralized budgets versus, you know, like why do you have so many engineers and analysts reporting to you versus like if it's kind of hidden in the business units right so it's kind of interesting how do you navigate those dynamics and every company is a little different in terms of how they look at those things. Yeah so, so quick question for what is your message right now to leaders so I see that the boards right so you made the case very specifically that the boards have the wrong mandate they're starting to make a shift but it's happening very slow so then there are these new jobs that you're saying they're coming up which are more now heads of VPs kind of so what is your message for leaders listening to us right now they're like what should you be like what should they be doing what should they not be doing where should be focusing where should they put their eggs in the back I mean what are your thoughts around this stuff to kind of give some recommendations. So I think this is very difficult but obviously the only I was having this conversation just before we came on air right so part of the issue is that most executive teams and board members they're not sat on LinkedIn reading my posts or your posts one right and I wish they were because that would be that would be good but they're not so it's very difficult to try and address this at any scale and I think that's always been the problem we've kind of been having the same conversations on podcast on LinkedIn forever and it's like well nothing's changed. You guys are in our own bubble too yeah this message isn't landing with the people that fundamentally make those decisions right so it's got to be changed from the inside out so and I think that's where the message to the leaders is that that's like that becomes the job I said something and I didn't think it was that controversial but evidently based on some of the comments on the LinkedIn post it was deemed to be which was if you go in and do you know as I said before if you deliver what you're asked to deliver then you're probably in a bit of trouble however it's the same like in the interview process if you try to challenge in an interview the likelihood of you getting that job is slim you know and I see this all the time right where really good people go in and they try to educate let's use the word educate they try to educate in the interview process to say actually you know you're saying that you need someone that's kind of come in and do this and this but actually we need to connect it back to the the volume levers of the business and what the business goals are and find what the KPIs are and usually that comes across as this person is trying to change the mandate to fit their skillset so therefore then they just get you know shoved off and they're you know left in the rejected pile so the controversial thing to say here is probably not along say what you need to say say yeah you know yes sir I can deliver all that for you and then when you actually get on the inside go and do what needs to be done so that's not to say don't deliver what they what they've actually to deliver but when you get on the inside you've got to try to change how they think about this because fundamentally if you don't and history has shown us you know time and time and time again that's why the tenure of the CDO is depending on where you read somewhere between 18 to 27 months there's a reason for that right and it's because when businesses go you know people go into these roles they deliver what they ask for they can't change the perception of what the role is they can't change the the kind of trajectory in the direction of travel away from it being just a building stuff building stuff building stuff and not connecting it to the business and the value levers then effectively they never get to the point of ROI I've seen the benefit and fundamentally it comes down to yeah it sounds really harsh but like do you want to keep your job on on and that's fine like you know a lot of people have made a load of money in this industry by being fired every two to three years and going into the same thing somewhere else and then the same thing somewhere else and the same thing somewhere else and then after five or six goes it they can retire happily ever after right and that's that's absolutely fine but fundamentally if we if we want to change the narrative of data analytics doesn't add any value then we've got to do something about it and it's really difficult as you know for someone like me I can sit here I can talk I can sell this stuff I can say this stuff to people in the boardroom that I do every single week it doesn't always land very well and they you know then choose not to do business with us because of it but fundamentally like we've got as an industry we need to try and evolve our thinking into how can we together collectively change this because that's the only way it happens at scale right okay oh man I got now we're getting even spicier but Tim you go first I got I got what do you have a follow-up to that well my meaning thought is you said this and how did people react to this particular because in a way what by the way I agree with you because and that may just be culturally how I how I was raised up on how I think about things and also kind of being a little bit of a bull in a china shop right you guys just cutting into the politics you got to do big friends and you got to understand and then like okay here's how I can go and and ask for forgiveness not for permission that type of stuff so that that's my way and maybe a little bit more of an American mindset probably but so yeah I'm all for that that's one thing second is is you you called it out people are like yeah I'm just gonna do a job other way I can do another job I'll do that I'll be comfortable and then like you know and and and and to be a little bit kind of I mean I mean the honest I'll be honest it's like you are part of the problem you are keeping this that is quo and you're not helping you're helping yourself but not the industry and fine you don't care about the industry I get that you can just get that that's fine right but I mean to be very direct on that stuff I'm actually gonna double I have to get this a little bit okay uh because like and it goes a little bit to what you said Kyle that like our ambitions should be greater but I think there's a lot of data leaders out there who kind of recognize that yeah I'm probably going to be like going three years because of who knows right it could be because of ROI it could be because of something else could be because new AI thing comes along whatever right but like if I could just leave things a little better than I then I let then I got here you know like then I did a good job and that's fine right and there's a bit of a like that's valuable but it's also pretty complacent complacent is the work here and I think this is this is this goes back into quote people who for listening. I'm doing air quotes over leaders, right? If you're complacent, that does not really, at least for me personally, a true leader is not somebody who's complacent. And I think this is where I wanna go challenge. If you are a true leader or you're a quote leader, and I think this is an opportunity to make a change in the industry. And you wanna be part of that transformation will change or you're just keeping the status quo, I don't know. - Yeah. - I think there's a couple of things on this one, right? So I think a lot of it comes back to skill set, right? Because obviously if you take this in its pure and it's never this kind of binary, it's not black and white. However, if you were to try to look at it through that lens, because of the types of mandates that boards create, you get a certain type of profile, right? And usually, not always, but usually that type of profile doesn't always come with the kind of prerequisite skill set to then be able to go and stand in front of a boardroom and say, well, actually we're doing this all wrong, and we should be doing it this way and this way because of X, Y, and Z. So all of the, you know, sales skills, if you wanna call them that communication, presentation, commercial acumen, you know, stakeholder management, influencing, educating, blah, blah, blah, often because of the mandate, because the mandate is, you know, we need someone that's great with, you know, building platforms and designing architecture and all that type of stuff. So there is definitely a kind of a skills gap and it's not always the fault of the person because if you think about it, and this is my, this is one of my kind of debating topics because what you find in the industry is that you, there's a certain group of people will turn around and say, yes, but it's the job of the data leader. And I'm absolutely like, it is, it is the job of the data leader. However, if boards keep appointing certain types of people and profiles based on what they want them to deliver and what they want them to deliver isn't aligned to the outcome that they are trying to seek then kind of what are they expecting to happen, right? So there's a bit of a skills thing here, but partly because they've never had the experience or the need to do it because they keep being asked, can you come in and build a platform, build, you know, design this architecture, build this framework and they do that and they get fired for it and they get asked to go and do the same thing somewhere else and so on and so forth. And I've said this quite publicly, like I think the way these organizations have hired for so long, it's taught an entire industry how to play the game in the wrong way, right? We're now at this kind of inflection point where everyone's saying there's no ROI and you're like, well, how do we fix that? That's not our problem, like we get brought into build stuff, right? So it's a really, it's a difficult, subject, but effectively, I think that the way those businesses have hired for so long, we've got this industry now that is like, you know, it's focused on the wrong things. But then also because of that, because it's never had the experience or the need to sharpen those skills or even use those skills inside most organizations, they sometimes don't have them, right? And therefore the ability to go in and change it from the inside out, it's easy to say, very difficult to do, especially if you've never had to do it before, or you know, never had the experience of doing it. And that's part of the problem. - So Kyle, what I wrote down here is like, if boards keep appointing profiles of people, of the outputs they want, delivery of platforms, frameworks, et cetera, instead of the outcomes they want, we're going to continue to see this problem. And here's my question to you, what should the mandate be? Like, what would be the ideal mandate coming from the board? And what's something we can do to start to shift them in that direction? - So I think, I don't think there is such thing as an ideal mandate as a collective because obviously it's massively context you will to the business and what they've done or had before and what they've not, and obviously sector, size, scale, complexity, or all plays apart. I think the one theme that should probably be threading through all of this is like, what are the goals and objectives of the business? Understanding what those KPIs and metrics are that influence those goals and objectives, and then understanding what the decisions are, that influence those KPIs and metrics that fundamentally lead up to the goals and objectives. And if businesses started at the end and work backwards as to what are we trying to achieve and how can data help us to do that to driving proof business performance, gain competitive advantage, whatever it is, then they would look at this through a completely different lens. And yes, of course, governance and architecture and engineering and analytics and data science, it's all massively important. You can't do any of this stuff without it, but in isolation, it gets you almost nothing, right? And I think it's that because the perception of what the role is leads them to create the mandate of going, building all this stuff, which is really useful stuff if something then happened with it, but it doesn't. So I think it's that if we need to get them to start thinking about this from the end point of what they would like to achieve within two, three years, and obviously business objective change fairly frequently, so it's hard to pinpoint, but they will generally be a handful of things that a board as a collective are trying to do. I had a conversation with the CEO, right? And he believes that he's gotten advantage because he reports to the CEO. And I asked him, what are you tasked with achieving? What are you deliverables? And he said, increase revenue and improve customer satisfaction. And he said, and the deliverables are the same for every single member of the board. So it was nothing about data or technology or platforms or frameworks. It was nothing about that. That he was just expected to do that stuff because it would help them to achieve as a collective group in the boardroom the goals and objectives. And I think when you frame it like that, the mandate changes into something that becomes a little bit more aligned to probably generating volume and ROI. So this is a lot with me. And I think that this should affect going back to what we talked about like the interview process, right? Instead of the interview being like, will tell us about how you approach governance frameworks and what's your ideal data lake architecture and yada yada, right? It should be, hey, our goals are to increase revenue and improve customer satisfaction. And then the candidate says, well, great. In my previous experience, whether it's in this industry or just relevant kind of experience, I implemented a machine learning approach that allowed us to take inventory from 90 days lead time to 60 days lead time. And I increased revenue by 5%. I can do that here. Yeah. In the takeaway, my main takeaway here-- so we have to go ahead and lay around quickly-- is one, ideally, we're going to have boards are going to start updating their mandates. So that's going to trickle down things that should get better. But in reality, this is where the true leaders are going to shine. Where they're going to figure out, hey, I really want to make a change. I'm going to look at some of the opportunity. It means I have to do a heck of a lot of work. I have to do all that delivery work, which I'm being asked to go do. And in addition to that, I got to play inside games for the inside to go change things. And that's what we have to go change, right? So it's double about a work right there. And that's my takeaway. So it's really for the data leaders. It's your time to really shine and earn that leadership and go drive that chain from the inside. And this is somewhat our territory. And let's go find our community and let's go do this together. Because hey, if you're listening to us, like we're here to support you, we want to see these changes. And there's people and businesses that have done this, that they are doing this right now. They are seeing tangible commercial returns year on year off doing this stuff. So go and seek them out. And I think that's one of the things about the data community is very open to sharing and coming together. So 100% I don't encourage people to do that. All right. Lightning round questions. Let's get this quickly here. And I know probably have more meetings this time today. OK, you know what's the funny thing? We haven't talked about AI that much. So actually question number one. Can AI kind of fix help the CDO Monday problem, or will it just make it worse? [SIGHS] That's a good question. At the minute, I think it's probably making it worse. Most of the mandates that I see are, you know, we want someone that's done Gen AI into production. And I'm like, OK, well, there's not many businesses that have done that, which means there's not for not many people that have done that. So I think it's making it worse. One of the terms could it make it better? It could make it better in the sense that if you view this through the lens of, if AI starts to be able to do more of the technical heavy lifting work, which we know it will do, then-- and I think that's going to be something that happens across the industry, right? Like if you think of, I don't know, analysts, are they going to be the ones that are building dashboards and reports in the future? Probably not. So a lot of that work comes back to QA that comes out of all of these agents and stuff like that, and more product management type skills and really custom-of-fate into the business and shopping in those skills, so possibly. All right. That wasn't lightening quick, I apologize. No, no, I was like, I like that. All right, second question. Is data-driven, which came up today, a mostly meaningless buzzword at this point? And how do you feel about AI ready, as a term? I think data driven is, yeah, it is a buzzword. I think it's always been a buzzword, right? I mean, how a business can be data driven, yet it doesn't know what the soon as it needs to make to impact the KPIs that it needs to influence to improve the goals and objectives of the business. So I think data driven was a better marketing thing, right, to be completely candid. Yeah. AI ready. I mean, it's like with all of these terms, right? There's logic behind them. You understand why people use them. They kind of make sense, but what was AI ready even mean, like to whom in what context? Like, you know, anyone's AI ready in theory, right? You know, like I can open Chuck G. PT and ask it something. Does that mean I may, I already, yeah, maybe, you know, I don't know. Like I think it's the quantification, the definition of this stuff is like hard, but I understand why businesses would use it. And they understand why data leaders would use it because, you know, if you say like data management, data governance, for example, that's long been underfunded, although some now they're getting loads of money thrown at them to do, you know, loads of work and projects that they couldn't get off the ground previously because they say, oh, you know, you want to be AI ready? Well, you better give me some money because data is really important, right? So, yeah, I think it's merit to it. I think smart definitions is key there, because what's funny, one quick comment is that they do surveys every year, you know, do you feel your organization is data driven, right? And it gets worse every year, like, like 10 years ago, it was like 35% now, it's like 20%. And it's because the definition's always shifting, right? Like we have more to do, it's getting more complicated. Yeah. All right, next question. Will AI agents make the CDRO more important or disrupted, distributed, displaced or? I think it will make the role-- I think it will make the role more important, but what I think is worth saying is that there's a definite land grab going more than minute. Like, depends on the organization who you speak to inside of business, is this a technology or is this a data thing, right? And I know many CDOs now who are-- they literally won't take a job unless they can have AI in the title. So now it's why there's a lot of chief data in AI officers and not chief data in analytics officers, right? So I think it depends, really. I think the answer to the question is, yes. It just depends on what the organization views AI to be. And if they view it as a technology thing, then the CIO is probably going to have the responsibility for it. All right, fourth question. If a chief data officer has a bad mandate from the board, can they define their own mandate and just shift to outcomes anyway? Yeah, yes, in theory. But as I said before, I think that's what you've-- I think you've got to get on the inside first and start to show what's possible. Because I think that's, as I said, if you try to do that before you get in role, and you try to do that throughout the interview process, like, yes, you can start nudging them in the right direction, but to fundamentally reshape what the mandate is. In my opinion, it's very, very tough to do whilst you're in interview. I think once you get on the inside, yes, you start to deliver what they ask for, but you also start to show them, eh, I've done something here. And look, we've managed to do X, Y, and Z, and show them what the outcomes could be. Then I think that will start to take them on that journey of actually maybe the mandate needs to be reshaped. But again, will that become an official thing probably not right? Because you just grow into what the mandate needs to be. And I think that's probably-- I'm going to say the easiest way, but the easiest way in that sense of changing it, right, from the inside out. Yeah. And then a couple of years later, they're like, man, Kyle's really valuable. Hopefully no one's ever said that about me, too. Kyle, you're really valuable. All right, take away time. Tim, take us away. Take away-- All right. I'll try to keep this tight because I know we had so many good takeaways today. So we started off with, what do you mean that people in the boardroom continue to design the wrong CDO mandate? And you said that fundamentally, however long that we've been doing data and analytics and LAI, there's a lot of talk and there's a lot of focus on technology, but not enough focus on the value and the impact and the outcomes. And a very finite number of businesses can really point to specific tangible value, maybe general value, but not specific tangible value around their data-oriented initiatives, right? And nine out of 10 times, the boardroom defines the role of the CDO around technology and frameworks and platforms. And now Jennie I is part of the picture, right? Rather than focusing it more on what are the goals of the business and aligning that role to delivering ROI. And the mandate is part of a bigger problem, the perception that what data means, that it's a technical problem that needs fixing and addressing, which is not what we want all the energy to be going into and ultimately results in the CDO getting fired into it three years, right? So is there a problem with the role of the CDO itself? You said, no, it's really around this mandate that needs to be changed here. And a lot of companies, they need to shift just focusing on things like, hey, let's become data driven to let's really make an impact on the KPIs that matter. And Juan asked, do we have more of a work problem versus the data or AI problem? And you said, you pretty much agree with that. And you tied it to a big problem around accountability, right? The perception of the CDO being more of an enabling or a supporting function versus really taking ownership over key things that matter to the business. And we need to switch hiring CDOs from more of a, hey, we're hiring them to do data and do enablement and instead to really impact those key things. There can't be this big disconnect between what is being asked of them and what's the outcome. We need to align those things. And you saw that there were some big shifts happening. The business is really looking at, now, and you can see this in the job descriptions or the interview process where they're signaling that, hey, we're looking for somebody with a strong focus on machine learning for pricing or something like that. This can be an indicator not necessarily that they're being technology focused, but actually they figured out the levers that are going to have a bigger impact on the business, which can also lead to things like shifting how you even organize this role to have bigger impact, right? But anyways, I have so many more, but Juan, I'll pass it to you. What were your big takeaways? - I love how we brought up a question. Why would a real CDO accept the role of the CDO or the real goals are not laid out? You're like, the goals are laid out, except that they're outputs and not outcomes. I think that's actually, if you're looking through the chat, we have a very kind of great chat. There's a lot of people who brought that up. You said nine out of 10, if you deliver what you were asked to deliver, you'll get fired because I was like, "Yeah, this is the unit, but it's costless." - Is it a set? - You don't need that. So your original goal were never linked to the outcomes. We talk about shifts, so you're seeing the shift, kind of the, you're always swing, but a shift right now to the centralization, you're having like many CDOs who are like the heads of data, there are been various parts of the business, and the roles are gonna become more specialized, right? So the data leaders want to kind of figure out, we're gonna put my eggs in what basket? We're talking about organizational structures and like all this data is born out of IT in the last five years, we've had a shift out of IT because they have not given the value, so they've moved to the CEO, the COO, the CFO, the two transformation officer, but the complaint is the same and we're seeing that shift back, right, pendulum always swings. And then people start saying, "Hey, why do we have like two CXOs "doing the same thing apparently?" And the CIO has more stock, so then the CDOs kind of get somewhat demoted or now you have heads up or VPs or now we're according to CIOs. But then we see kind of this matrix organization where there may be a point to the business line, but they're dotted line centralities as some of the IT functions. Recommendations is like, look, we're having this conversation, but this conversation isn't always landing to the exact and the point because hey, they're not following us on LinkedIn and listening to our podcast, but they should. So we need to really do this from the inside, the data there is need to do this from the inside. So hey, if you go to the interview process and if you're the person who's challenging or educating the interviewer, it gets perceived that you're trying to change that mandate. So that doesn't work and you're suggesting which can be controversial, but I agree is just not along. So yes, I can go do that and then get inside and then change the things from the inside. A lot of it comes back to the skill sets, the type of mandate boards result in certain types of data leaders. So what Tim was saying, if the board keeps appointing profiles of people of the outputs they want, instead of the outcomes they want, we'll continue to see this issue right now. And the ideal mandate is not universal because it's contextual to the industry, but one common thread is, what are the goals and objectives of the business and what are the KPIs that influence it? So a CDO that reports the CEO, my goals are to increase revenue and improve customer satisfaction. Yes, that's the type of somebody you go focus on. How did we do? Anything we missed? - I don't think so. I don't realize I said all that, but yeah. - Well, thank you. - What are your teachers that came out of that, I think? - Yes. What's your advice? You did a lot, but what's your advice about data, about life, whatever you want? Second, who should we invite next? And third, what resources do you follow? - What's my advice? Well, advice about life is, you know, be happy and enjoy it because we're not here for a good time, not a long time. About data, try to focus on outcomes, try to just try to drive that message home, that's the only way we change anything. Who should you invite next? This is really difficult because I know so many great people, I'm trying to think as somebody who will be, as, you know, honest, no BS. Have a going talk to Davin Crowley. sweet who is this chief data officer at National Highways. So he's somebody that's done phenomenal work in this space. He's been given an OBE here in the UK for his work and he's got some fairly strong opinions. He comes from a music background of all things. So yeah, I get him on because he loves some pre-controversial views. And then resources, I read. Sorry, there's some things. All right. I read a lot of books, but to be honest with you, most of my stuff comes out of LinkedIn and articles and stuff like that. So yeah, I don't say there's one main place. I think to be honest with you, where I get most of my knowledge is by interviewing people on my podcast, right? That's probably why I'm not. That accompanied with interviewing people for CDO roles and speaking to boards about the mandates of CDOs, those things collectively are where I get the most insight from like real life on the ground stuff. Awesome. Kyle, thank you so much for for finally being on the podcast. I'd love that we have these conversations. We can keep going on. I look forward to meeting you in person. Let's hang out and have some whiskey and some old passion, some Belgian fears. Yeah, let's do it. You know, that keep happens conversation alive because we need to have that and get some board members too in the room and we can have this. 100% years. Thank you so much. Bye. Bye.