Episode 4: Steve Green

AI Transcription:

Ben  0:02  

Hi. This podcast is brought to you by Practicus. Practicus is a recruitment, consulting and advisory business, specializing in change and transformation. We hope you enjoy the podcast. 

Tony Cassin-Scott  0:13  

Welcome to The Data Files data bunker. Our next podcast again is on data leadership. Today we have myself, Tony Cassin-Scott, 

Barry Panayi  0:22  

And me, Barry. Hello, 

Steve Green  0:24  

And me, Steve Green. 

Tony Cassin-Scott  0:25  

Thank you for coming, Steve. So Steve has a very interesting and broad background, both IT and data, and recently belive Steve you were the Head of Data at the FCA? 

Steve Green  0:37  

Yep, I was there for planning that data function across all disciplines for about four years. 

Tony Cassin-Scott  0:41  

Wow.

Barry Panayi  0:42  

Thanks for coming here. Honestly, we do appreciate it. I think having someone who’s been on the regulatory side and the commercial side is pretty rare. And then to get someone who’s done that with IT and data is even rarer. So I do appreciate you coming down to the bunker. The question we’d like to start off with is, what do you think makes a great data leader?

Steve Green  1:03  

I think it’s a combination of skills. You’re always going to say that, aren’t you. You need to be able to understand the business enough to understand how you get value from the data. I think classically, data leaders were seen as being predominantly focused on data governance and data quality. And whilst that’s clearly important, and particularly from a regulator perspective, I would say that, historically, actually, the excitement you want to generate in your business to really change the way the business sees data requires the data leader to be able to sit in front of business strategists, the executive board, and talk to them about actually, as part of your strategy, this is how data can help you. Not be at the end of the process, just talking about how you might write a new dashboard or write new analytics project. And so that combination of business skills, clearly understanding both the technology and the techniques that are available as well, you know, so you can have that end to end conversation. But the empathy around understanding the business problems, how do you solve them, focusing on outcomes. 

And then clearly, you need to build a good team. You know, data is a critical resource at the moment. And you know, London, there’s 10s of 1000s of roles you hear about. And the idea is, actually you’ve got to retain your staff. So you got to be careful that you don’t just hire the best people, leave them alone, and then they get bored, and then leave. So that combination of business understanding, being able to deliver business value, understanding the technology and the techniques that support actually the delivery in your function, but then also being able to care enough about the team, so that you can keep them with you. Because whilst it’s an exciting space to work, that also means it’s exciting to work elsewhere. And so there’s lots of opportunity there to really make sure you can hone your team into a good unit who feel valuable because you expose them to business strategy, they can see line of sight to how they’re delivering stuff better. But that requires a leader who can do a little bit of everything. I think, clearly you can be a good data leader if you’re focused in the data science space, or focused in the analytics space, or even focused in the technology space. But you need to bring that with you something wider than that, that ability to influence the business and deal with the business needs.

Barry Panayi  2:51  

There’s a lot of stuff there. So a lot of good stuff. I’m just picking my way through where shall I start? You intimated a deep knowledge of data in some in some way. That seems to me like tickets for the game. I’m leading the witness a bit. But do you believe it’s possible to be a CDO or equivalent without ever had got your hands dirty? I declare my interest.

Steve Green  3:17  

So you know, so I am somebody who historically used to code in VB and basic stuff like that. So I’m, I’m somebody who is reasonably technical, I would argue, but I’m a history graduate. I’m not STEM graduate. And actually, I do use this as a bit of an anecdote some of the teams when we’re looking at who we want to hire and not exposing ourselves just to STEM graduates, although the clearly there the core of the team. Is that history for me at university was reading 1000s of pages of documents and concisely producing them into six page reports. So from a middle management and senior management perspective, actually, the skills history taught me around digesting data, information, just not numbers, and binaries and stuff like that, actually has helped me in the mindset from an analytics perspective. And clearly, I’m not going to go do some data science projects, although I understand the techniques, I understand the projects and the value they can give you. I’ve spent 25 years in around IT and data. So whilst my background is not STEM, clearly, I’ve picked lots of stuff up. So I think it’s a balance, I don’t think you need to be the hardest data scientists or the best data scientists to run a CDO function. I think you need to clearly understand the possibilities and the opportunities and you need to understand what interests your staff as well. So you need to understand the techniques and technologies and you know, where the where the cool stuff is. But you need to be able to do that in a window of not just been the technologist in the room, you’re not the techy person in the room, that’s your team. You’re the person who understands the techie stuff and the data stuff, and can influence the business to using it more effectively.

Tony Cassin-Scott  4:41  

You mentioned a lot about leading a team and recruiting the right team. But beyond that, I would argue a data leader also has to get buy in.

Steve Green  4:51  

Absolutely. And that’s that engagement with the strategists in your organization. So particularly, you know, experience both at the FCA and my current role at Sumitomo Mitsui, you’ve got to spend a lot of time pressing the flesh and talking to people and not just talking to them after they’ve made some decisions, you’ve got to talk about, clearly there’s a minimum viable thing you need around data governance. And everybody knows that, you know, almost every organization in the country will say, Yes, I have to invest in data, because legally, I’m obliged to do so. But if you want to really excite them, and to get you involved in all of those big conversations that you really should be involved in, I think, you have to get to know people, you have to build their trust, you have to talk to them in a way that is not about natural language processing, or AI, or this algorithm or this model, or even things like data ethics, you want to talk to them about so what’s your business problem, I think I’ve got some ideas on how to do that. Deploying quick value kind of solutions at the strategic level, you know, dealing with senior directors and executive members who can then go, actually, this is quite cool, we should care about it. And so that, to me, is not that dissimilar to how IT evolved about 20 years ago. I mean, IT 20 years ago was build it and they will come and not a lot of people came except where they had to. And so in a world where data was where IT has moved into being top table type conversations where you need to care strategically in your business around IT. Businesses say they do that with data at the moment, but a lot of businesses still only care about it from a compliance and risk perspective. It’s when they when your CDO or your head of data, whatever it might be, can go into the business and go, I can fix your problems, not just fix your compliance risk. That’s really where the conversation needs to be. And at that point, you do influence the leaders because you’re talking their language, you’re talking their value. You’re not just talking about we need 12 data scientists and I need to lock them in a basement do cool stuff. Because that doesn’t resonate. You need to be able to talk to them about, you need to increase revenue, you need to spot opportunities in the market better. You need to manage your risk better. Those are the conversations people want to have and all the CXOs care about that.

Tony Cassin-Scott  6:47  

So over time, have you seen a change in the in the board mentality? So you spoke about exciting, the sponsors, the board, etc? to get on board? But the board? Yeah. Are you seeing it the other way? Where there’s that’s a push? Are you seeing a pull now? Where people get it more than they did before. And then they’re asking you for things, knowing what’s possible and what isn’t?

Steve Green  7:08  

I think there’s a bit of that. I think there’s still a large gap between you’ve got people who say I need to care, I know I need to care because of risk and compliance. So you know, that’s, that’s the easy stuff. You’ve also got people who think you can solve everything. So they think actually throwing those 12 data scientists into a room will get you really cool outcomes. Without the business analysts without understanding the business without working with the management teams in line. A successful delivery, again, I get back to my experience in IT 20 years ago, IT thought hiring all the best developers and techies would give you a great delivery project, loads those projects, overran cost 10 times the amount of money, whatever it might be, because they didn’t understand the business, the real outcome they were trying to achieve. And so if you can get to that with your team that has these other skills in it, then I think you will certainly make it a huge difference the organization.

Barry Panayi  7:55  

So there’s a certain level of leadership skills, maturity, you’re intimating, when you talk about those things, I’ll go out on a limb and suggest a lot of very good data people don’t have those skills. And do you think that’s what’s maybe stopping the CDO or equivalent getting traction at the exec or board table? We don’t see CDOs becoming CEOs or even COOs or even sitting on the Exec? And is that the gap? Do you think and it’s not the fault of us data people not thinking it’s important? Or has the organization not developed? Its data leaders like it would it’s finance leaders or people leaders?

Steve Green  8:39  

I mean, clearly, it’s a newer discipline. So it always take time, I remember 10 years ago that most of the CIOs I’d worked for, had never actually done on the ground IT roles. They’d been brought in from operations roles, and therefore, they never actually understood what made an IT department tick. And so they were finding it really difficult to make their department successful. I think in the database, it’s gone the other way where most data staff have come up through the ranks of data functions, whether it’s analytics, or data science, or data governance. And you think you’re right, the exposure to business strategy just doesn’t exist that much. And even three or four years ago, the vast majority of CDOs, I would speak to at conferences, and I’ve done hundreds of conferences over the last five or six years, we’re predominantly focused on either you are the analytics person, or you are the governance person. Not you are the leader within the organization, you sit on the senior leadership team, your job is to help influence strategy. So actually, that career path and that learning is really difficult, because there’s no steps to make, you know, you either you are a data governance, you are a data analytics, you are this. And then suddenly, you have to step up to being this strategic thinker who cares about the entire business. And it’s how do we find that middle ground? How do we as data leaders expose our staff to those business problems, not just data problems?

Tony Cassin-Scott  9:53  

And do you think that’s a function of non data management or data management? What I mean by that is, should the day people go and find their own path to success? Or is it beholden on the rest of the organization to understand and value, the skills that the data people have and actually grow those skills that can be more helpful to the rest of the business, which is it was a both.

Steve Green  10:15  

So it is a bit of both, I think if your organization doesn’t understand what’s possible, they’re not gonna care. So as a data leader, you need to inspire them, you need to show them what can be achieved if they treat the data functions as a more of a strategic asset for the organization, front office, middle office, back office, whatever office it may be. And I think if you can get to that stage, then your business and your HR teams and your career paths and your center of excellence and whatever it might be, will be cared about more deeply. And therefore people will think about leadership skills within your data functions. If at the moment, your data office is only doing governance, and you know, and your analytics team is doing analytics, they will box you, you will be in a hole of not just your own creation, the organization’s creation. But you should be able to step away from that by looking up. And I’ve noticed over the last two years at conferences, more people will talk about the first conversation being about the business problem now, three or four years ago, you hardly ever saw that it was about how do you hire data scientists, How do you stand up teams, And how do you do this? All really important. But that business change, work that BA level work, business analysts work that engaging as say with the strategy while it’s been written, not after it’s written, and you’re being asked, Can you help. IT have moved through that chain over the last 10 years? You know, IT teams will often be at the point of strategies being built now because you have to understand the capabilities available. I think in most organizations, the data seems not there yet. And should be. 

Tony Cassin-Scott  11:32  

If data teams are following like the IT lead, as it were. Do you think that’s a good strategy? Because I would argue that I wouldn’t say that IT departments themselves are kind of leading business change.

Barry Panayi  11:45  

I’ve been slagging them off for the last few podcasts.

Tony Cassin-Scott  11:47  

You have indeed.

Steve Green  11:48  

No, and I think there’s I think there’s a big difference between following the path IT have gone. And I would argue that in many organizations, IT has not got to where it needs to be in that path. There’s a reason why Agile is mixed value teams. It’s got business staff, and it’s got developers and it’s got all the other stuff that you need to run an agile kind of team. In the data space, you need to have business analysts need to have business people in your projects. You don’t just want to have the data scientists saying bluntly, occasionally, arrogantly this is what you need from us. And you don’t understand and why do you still work that way? It’s all a bit rubbish. And I’ve seen data teams do that. And I saw IT teams do that 20 years ago. So it’s more about the journeys, similar around, understanding you are there to work with the business not to dictate to the business from a center of excellence. But you’re right, you know, my last two data roles have not worked into IT. One worked into the innovation division in the FCA, because its job was to change the FCA. And at the moment, I put into effect the the kind of COO level roles. So it’s not a technology role, you know, and I would argue it shouldn’t be just a technology function. So maybe the you know the name, IT is obviously a bit of a misnomer at times. If you’re talking about that kind of IT division, then yeah, it’s definitely very different. I think that actually gets you to an IT division that’s very similar to just a data governance function, where you’ll you’re providing a base service that people know they need, but they don’t really love. And I think what you need to do is create that engagement of the business again, where your data team is seen as not just being a data team. It’s full of problem solvers and stuff like that, and people who will listen to the business and understand them. So again, the IT analogy is more about the path that I’ve seen IT following last 20-25 years, as opposed to us necessarily treading exactly the same path.

Barry Panayi  13:25  

It seems to me that to be a data leader, you need to be a leader, business leader strategy conversations, looking at the teams you’ve managed and the people that you’ve met, we talk about data scientists and data engineers being this commodity, paying through the nose to get them in. Can they be arsed to become a CDO or head of data? Because where is the pipeline of CDOs coming from? Is there such a rampant market for technical specialists, that perhaps it’s old fashioned to expect them to want to grow up and be CDOs? I mean, I’d be interested in your experience, have you got a succession plan where there’s a hungry Data Director or head of data science who can’t wait to be in your shoes? Or do they think God that’s a ball like politics, Board Papers, and so on?

Steve Green  14:17  

I think so. I think you’re in a situation where there’s always people who enjoy the problem solving, rather than necessarily the technology or the techniques to get there. And if you can expose them to those business conversations, rather than just giving them a piece of work and going please come up with a model or whatever it might be, if they can be exposed to the business benefit and the fact they can really see themselves moving the dial in the way the business works. And they can then appreciate maybe more closely the feeling of value the business then has for them as a function. The more senior you get, the more of those things you can do. You do through your teams, but you can fix more problems. You can deal with more stuff, you can make more of a difference. The same conversation occurs in any technical discipline, whether it’s law, accountancy or in it, if you’d like Like the day job, some people have to step away from a bit of the detail to become the management functions around it. And that will happen everywhere. Again, I think because data is a new discipline, it’s not really new, you know, analyst institutions have been around for years and centuries. But you know what I mean, in the current guys, it’s kind of a newer discipline, I just think we’ve had less people have the time to make that journey. And I think as you get to a place where more people are in, you know, in my bracket in their 40s, or whatever, and they’ve been exposed to 20 years of data projects, shall we say, or data modeling, at some point, someone’s going to want to do something different with their experience. And so I think we will see people then rise into those functions, what you’ve then got to do is make sure that you didn’t treat your team just as a technical delivery function. Again, kind of learning the lesson lessons of bad it functions where the CTO can never do strategy, because they’ve only ever been a technologist. Actually, that’s not what people want. And that’s not what people should want from the data functions. I think too often people have been pigeon holed, and therefore they’re not encouraged to look up and look out. I think the best kind of data functions in the country now will be those that say, Actually looking up and look at is exciting. It’s interesting, you get to bring all your skills, you get to use them in those meetings, where you inspire the CEO and say, Actually, I can do this and this and this, you know, when when we’re when we’re the FCA, we had kind of one particular project that we worked on, which really inspired the board to to invest heavily in what we were doing. If you can give a data scientist that exposure, and then say to them, actually, you can be doing this a lot of the time, if you come up through the management layers, a lot of them will that would accept that I think, plus, you also get the added benefit, which is one of the reasons I moved to management is that then have to keep my skills up to date in quite the same way.

Tony Cassin-Scott  16:42  

Quite right. Well, I’m endorsing that. For what you describe, it’s a function of, of time and maturity going hand in hand. In that, if we were having this discussion and 10 years time, what would be the content?

Steve Green  16:53  

It’s an interesting journey. If you get to a space where the data leaders are able to influence business strategy and therefore maybe lead aspects of business transformation, they then get exposed to the end to end view of business change. Once you’ve been exposed to that, there’s no obvious reason why a coup well wouldn’t be achievable by the CDA, because they’ve got all the experience you wouldn’t need. Because you know, if you are a data led business, then your change should be data led, which means your current state should be data led, which means you are data led, and therefore, the CDO should be able to support that. I won’t quite let people off the hook and say it’s only because of time, I think leadership in organizations does not yet appreciate the opportunities available to it. And therefore, they easily focus on data governance, because the regulators of the EU will say, you must care about data governance. And nobody the EU is saying you must build great analytics, you know, no one’s in that space yet. I think the leadership teams need to use all their resources to the best they can. And in many organizations, they are not using their data teams to the best they can. They are pigeonholing them in risk, compliance, visualization only. They’re not asking them to the top table, which allows you to really change the way your business is. It’s a bit like automation. Really, if you want to do automation properly, don’t just automate your current process. You have the automation teams in your design meetings around why are you here as an organization as a department? What do you really want to achieve? And then you work out how automation can fix your outcome achieve your outcome, not just how it automates three or four manual steps. That’s how you make real change. And too many leaders above the data functions are not having those conversations with their teams yet.

Tony Cassin-Scott  18:28  

What will it take for those conversations to take place?

Steve Green  18:31  

inspiration, perspiration, all that kind of good stuff. It is partly the fact that the data leaders need to drive those conversations. As you see more examples in industry, I think that’ll be good. It’s always an interesting one, when you look at kind of top data leaders, and you know, most of them are either consultants or they’ve come from like a data governance background. And they’ve, they’re the CEO in such in such an organization, that bluntly, most of those CDOs are data management people. And if your scope of conversation is purely data management, then you’re never going to get out that box. Because you have to show them the value that good data quality, you know, you aren’t going to automate any process if your data quality is not very good. And so actually the data governance conversation should not be about DQ for DQ sake, it should be actually we’re enabling you to automate your business. That’s a whole different conversation from I need to care about hitting 90% data quality in this field. So again, it’s the tone, it’s the change, it’s it’s all of our job to kind of change the narrative a little bit about data leaders, I think.

Tony Cassin-Scott  19:27  

But there’s, there’s two sides of this. I think the first point you made is if we’re coming from predominantly a data governance perspective, that quality perspective, and that’s what the conversation is. It doesn’t move the dial that emotionally. Secondly, though, the people receiving this information, I mean, it’s not like this is a secret society. We all do sort of data analytics on qt is widely known. We’re discussing it now. Isn’t there a onus on those people, the non data people on the board to actually get more out of the data function?

Steve Green  19:58  

I think that’s what I was saying earlier. You know, the board has a responsibility to use its resources effectively. If it doesn’t understand what’s possible from those resources, then it’s not going to do that. So they have to spend time to get to know what’s possible. Now, again, that can be done through the data functions, you can do leadership training, you can have many different firms will give you their academy training across different levels of your organization. I think if you can do that, then that works, because then that’s a little bit of an education before they then are expected to pull. But But again, it’s very difficult for people to know what’s possible when they don’t know what’s possible, you know, if they’ve never come across it, they’ve never worked in an organization that’s run that way. They’ve never worked in an eBay and Amazon. Actually, most organizations have lots of data, they do MI, but they’re not really kind of exploiting their data to the best of what you’d expect them to do. And I think it just, that will take time. But again, open the door, shake the hands meet the board’s as a CDO, get out there and talk to people. And don’t just talk to him about risk and compliance talks about what the outcomes you can achieve for the organization. 

Tony Cassin-Scott  20:58  

I think some of the barrier is it’s a bit of a black box it’s a dark art. It’s too hard. So it’s very difficult to get that conversation going, when you’re kind of hobbled with with that prejudice. 

Steve Green  21:11  

And I think that comes down to the language use. And I think that’s probably the weakest thing in the data conferences at the moment is there’s very little focus on that kind of outcome strategy language with the organization No, and lots of data strategies do just focus on how do I hire more data scientists head or do my analytics, how to do better data management? Where’s your data strategy should be the language of how do I grow revenue? How do I spot opportunities in the market? How do I do this and do that? So again, it’s it’s a really difficult thing to do without experience across both sides. And there’s not many people who’ve got that. But we as an industry need to change the language that we are using those conversations, because what’s behind the curtain might be blind to the organization. But the way you present that as a lead, the way you talk about it, and it doesn’t need to be a CDO, or your your management team talk about it should all be about why you’re here as a business. What are your problems? What are the risks you got? What are the things preventing you achieving your revenue targets or your outcomes, whatever it might be? Okay, so what information do you need to spot prove disprove scale evidence, whatever your problem might be? At that point, that’s a business language conversation you can talk about, I wish I could see what’s happening on the high street in Bradford or whatever, that is not a technical or a data conversation, but your business leaders will go, actually, I need that intelligence, I understand it, then you can go and have a conversation about so what data do I need to acquire? How do I give them that data? How do I use it and analytics or visualization, whatever it might be? Some of that conversation can be behind the curtain. But you got to expose the language in a way that makes sense because it is unfair unexpecting, a CEO, a COO, to understand all the stuff you know about everything you know about data techniques, and tooling and stuff like that.

Barry Panayi  22:49  

Absolutely. And if I may go slightly more controversial, but I think you’ve you’ve touched on it, actually, in terms of the language and the way we may focus on governance, data management, we’ve spoken a lot about what makes a good data leader, and that you’ve painted a really nice picture there. What traits of bad data leadership have you seen out there? You’ve been on the circuit and probably heard the same people I’ve heard year after year and had opinions we won’t name names, because it presumably we’ve nobbled ourselves as an industry, because we’re finding our way a bit, you know, it’s not been a straight line. Yeah, what what have you seen out there that you think we should be avoiding?

Steve Green  23:25  

I think the biggest mistake I see is arrogance, because it’s the cool thing is that you can’t be the person dictating to your business, how they should work, you know, we may be the people who have some sight of what the future might look like. But that doesn’t mean that everything that’s gone before is rubbish doesn’t mean that people don’t know anything, it doesn’t mean that you should wait and go. Now I’m here to save you all from yourselves. And I’ve seen people do that. And I’ve seen conferences talk like that, where they talk about, you know, the future is data. And that kind of dismisses the input that’s required from the people running the current process, who’re making the sales at the front door, all that stuff is required to be successful in the data enabled business. And so I think the arrogance is a big thing. I think also just focusing on the skills, so everybody knows they need to hire, and they know that hiring is difficult. And so if data leaders, just focus on getting the right people through the door, but then don’t think about your path to life, are those data science products? You know, how do you actually make them real across the enterprise? How do you sell them across the enterprise? What’s your comms approach? How do you empathize with the business? So the narrative is not your narrative, but their narrative? There’s just a whole bunch of again, it comes back to having the right skills, but you know, classic T shape, isn’t it rather than I shape, I think to be the leader in most spaces nowadays. But the worst thing I find is where we’ve had to call out people for being arrogant about their capabilities. And just winding people up on that basis. And I’ve seen conference speakers, jump onto stage and talk about how amazing everything is and not appreciate that actually their new process their new product, their new thing comes from a mixture of the data teams and the IT teams and the business teams, you know, the kind of front office teams, whatever it might be. And I think you, we have to break that down just because we’re the shiny new thing doesn’t mean we are the best thing.

Barry Panayi  25:16  

Here here. What a fantastic way to almost finish the podcast. Unfortunately for you, we always finish these podcasts with the same question, can you tell us what’s the best advice you never took?

Steve Green  25:31  

Probably the best advice I never took is I was in one role where I leapt too quickly because I was bored. And didn’t really consider that staying in one place sometimes allows you to go up next. And so I’ve spent 20 years, give or take when and get the kind of middle manager layer. Because I moved around lots because I was following the interesting work. That’s been a really good thing. But it has meant that there’s probably levels of there’s probably been roles that I’ve not been able to get because people have seen me as just that kind of constant consultant who runs around that middle management layer. And unfortunately, you know, my work in the FCA kind of allowed me to step up into a different type of role. And that’s allowed me to expose more people and more things to how I think and how I work, which has been better for me and better hopefully, for them. We’ll see. And so I think that, that pushing too hard for the next shiny thing, as opposed to taking a slightly more considered view of where you might want to be in five or six years time. I think I probably made a couple of mistakes there in my mid 20s. When I was like, actually I’m just a bit bored. I should move on. I don’t really regret it. But I can see how that might have changed slightly my career path over the last not so much the last five years, but certainly during my 30s.

Tony Cassin-Scott  26:43  

Thank you very much. They’ve that’s valuable, incredibly honest. Thank you. Yes. And unfortunately for us, we do have to end it now. But thank you again Steve. 

Barry Panayi  26:51  

Thank you very much. 

Steve Green  26:52  

Thank you.