“Treating data as an asset for me is about treating it in the same way that you would other assets that are maybe more readily recognised as assets… It holds realised value, probable value, and potential economic value to the firm.” – Denholm Hesse
In this enlightening episode of the Digital Lighthouse, we sit down with Denholm Hesse, Chief Data Officer at BDO.
With a rich background spanning over a decade in data management, including pioneering roles at Deloitte and Prudential, Denholm brings unparalleled insights into leveraging data as a strategic asset.
Our conversation dives into the evolving role of data governance, the innovative concept of a data marketplace, and the critical importance of data quality and trust in driving business efficiency.
Join us as Denholm shares his visionary approach to harnessing data’s potential in turbulent times, illuminating the path for businesses aiming to thrive in the information economy.
Transcript
Zoe Cunningham:
Hello and welcome to the Digital Lighthouse. I’m Zoe Cunningham. On the Digital Lighthouse, we get inspiration from tech leaders to help us shine a light through turbulent times.
We believe that if you have a lighthouse, you can harness the power of the storm. Today, I’m excited to welcome Denholm Hesse, who is Chief Data Officer at BDO.
So hello, Denholm.
Denholm Hesse:
Hello, thank you very much for having me.
Welcome to the Digital Lighthouse. Tell me briefly about your background and your role at the moment, like what you’re responsible for.
Denholm Hesse:
So, I’m the Chief Data Officer at BDO, as you said. I started my career in data around 10 years ago, working at Deloitte in a consulting capacity. I was part of Deloitte Digital, actually, in its very early stages, which was a lot of fun. I started as an analyst building Tableau dashboards. It was a period when putting a map on a dashboard was really exciting for people, but very tactical bits, and I was doing it pretty badly, to be honest with you.
Then I started to transition into managing data projects and data deliveries. I really enjoyed that. It taught me a lot about how to manage people in technical roles, how to explain technical concepts to senior stakeholders, and how to work with clients.
But it also taught me how to get things done, how to get deliveries over the line, and I think that’s a very important skill. I moved from Deloitte, then to Prudential. I was actually a consultant at Prudential with Deloitte; it was a seamless transition for me, but I moved in as the head of data. The job at the start was essentially to stand up a new data analytics team.
So completely Greenfield, built a new team from the ground up to about 10 initially, and over the four and a half years that I was there, it just grew. Prudential merged with MMG on the FTSE 100. By the time I finished up, I was director of data that had a team of 75, across data engineering, and analytics… digital analytics.
We delivered analytics globally, basically, for the client-facing and distribution out to the business. Then just over a year ago – so Jan, 2023 – I started my new role as Chief Data Officer at BDO. BDO, for context, is a, you know, we’re on track to be a billion-pound business. We’ve got eight hundred, 1000 people working for us in the UK. And we provide audit, tax, and advisory services, the market leaders in the mid-market that’s going to attend to 75 million pound revenue firms.
And I guess what really drew me to the role was one, BDO has got a really impressive track record of growth, even through difficult conditions and difficult markets. They invest really heavily in people; they know that people are kind of the core of the business. But also, the role itself was what I consider to be kind of a true CDO role. It has the full mandate, which is very important to me.
So I look after data management and data governance. I look after data analytics and engineering and look after data culture for the firm. And I think – before we get into the topic of the podcast – I think the thing that I’m enjoying the most is the kind of shift from the head of and director role to the CTO role, which is for me shifting from running a technical team or technical delivery team or teams to being responsible for data as a whole and its use and its misuse, I suppose, across the firm, inside or outside of your direct teams. So are we making the most out of our data as a firm, regardless of whether it’s my direct reports doing that, my role is now data right? Across BDO. So I think that’s what the purpose is see into the CDO title, it makes it a very big task, but enjoying it so far.
Zoe:
A very big task and very broad. So I think it really sets the theme to look at your background and where you’ve come from, and all the different facets that make up this role that you’re now doing. But I’d quite like to ask about the core at the core of it, right, is what data BDO has, and then how you utilise that as an asset. So, you know, what are the kind of tangible outcomes of treating the data that BDO has as an asset?
Denholm:
So it’s a good question. I think so for me, when we think about data as an asset, you know, we think about lots of firms will say, well, data is really important to us. And I think that’s great. It’s a good starting point. But treating data as an asset for me is about treating it in the same way that you would other assets that are maybe more readily recognised as assets.
So maybe property or real estate is a good example.
Zoe:
Right? Like physical assets.
Denholm:
Yeah, exactly. Yeah, more tangible assets are the easy ones to recognise. So when we think about data as an asset, we think about it in the sense that it holds realised value, probable value, and potential economic value I’m talking about here to the firm, and that it’s identifiable and it’s separable from other assets.
So for example, data assets are distinct and separable from our technology assets, I can tell you what they are, I can identify them. And I believe that regardless of the technology that they’re stored in, they hold some probable or potential economic value to the firm.
So we think about it in a real like, you know, and working accountants now we’re thinking about it in terms of how do we define an asset. So that’s how we think about it. So it’s more than language, I suppose in terms of data is important to us. We want to be data-driven, that kind of stuff that you’ll typically see in vision statements.
And I think what that then says, when you start to think about it and get it recognised as a genuine asset, as you say, Well, okay, that means we should manage it.
And we should expect a return on it, you know, we should recognise that probable and potential economic value. And so you say, Well, how much money and effort and time should we spend on that? And then you start to get into the question of, well, how much is the asset worth? So what is that potential value? What is that probable value? And I think that’s a really nice kind of shift in mentality for the business to start thinking about data in a different way.
Well, also, money is the way that we can compare resources across different parts of the business, right? What do we invest in consultancy versus invest in data?
Well, actually, it’s totally different parts of the business. And money helps you turn it into a single dimension that you can compare on. And I think this is a great way of looking at it to say, what return can we expect that both sets the expectations because I think it’s too easy to say we value data, that actually you leave it sitting around, and you don’t make the best use of it. Because it’s not in anyone’s priority.
But actually, if you say this asset should be generating this revenue, it kind of drives you to make the most of it, right?
Yep, it’s useful for, you know, lots of different groups of stakeholders for analytics teams, you can look at it and go Well, which are the highest value, data assets. And really, that’s a good starting point. So what maybe that’s the place to explore what data products we can build with it. And that might be reporting an EMI, which is kind of the really traditional use of data, but it might be new revenue streams, but is there an option to monetize that data out to the market, for example, because it’s that valuable.
So you can kind of change your perspective and maybe make the data team feel a little bit less inward and think about the value that it creates for the business. It helps you when you go and have conversations with people within the business within their role. So we’re collecting data and putting it into the system because they start to realise its value to the firm. And one of the big challenges for data governance teams is you go to assign a data owner or we call them data trustees, but data owners what is more typically known as, and you find somebody who’s relatively senior in the firm, and you say you are accountable for the quality of this data, and how it’s defined, and its use and all that kind of stuff.
And they’re really busy people who haven’t worked in data teams specifically. So maybe don’t feel comfortable with all the language that’s been used around there. And now they’ve suddenly got all these new responsibilities, and they got to try and galvanise a lot of business areas to get behind them on something.
When you start to talk about them, you’re raining, I’ll make it up, but a million-pound asset here, it changes the conversation again, and it helps them you know, even in their own kind of personal career development, when they’re doing their performance reviews and things they can say, well, I’m responsible for this million-pound asset in the same way that someone who looks after property and real estate will say I’m responsible for this, this size asset, he can do the same for data. So yeah, it’s a good sort of shift, I think, in perspective, but it also helps you kind of get practical in terms of how you sort of allocate your effort to managing data across the firm.
Zoe:
Yeah, and it’s obviously going to change on a tactical basis. You know, exactly how people are treating it day to day. Does it also change it at a strategic level in terms of decision-making?
Denholm:
Yeah, for sure. Because I think there’s kind of two parts to it. For me when I think about strategic decision-making. One is you actually have clarity, and you have trust around data to bring it into that strategic decision-making kind of sessions and meetings and wherever that might be.
So you got more confidence and having data as part of that process? Secondly, you know, what are the businesses that win and do the best, they make decisions better than every other firm, all their other competitors.
And so data actually, by managing it properly, and making sure that it’s democratised, it’s accessible, and it’s well managed and maintained, means you can use data to make decisions more frequently. So it’s about “Yeah, I can trust it.” And the other one is about “I’ve got it there for me and it means rather than it being a yearly cycle, I can look at this across the year, and we can be genuinely Agile to what’s happening in the market”, or, you know, the stuff that our customers are sharing back with us. So I think those two ways are probably the two thing is our linkage to the strategic place?
Zoe:
Yeah, that’s super interesting because my background is in technology rather than in data. And so it seems like there’s a really clear parallel to me with this idea of “once you can release your technology frequently, it changes how you can use that and what you can expect from that whole function”. And obviously, doing a piece of data analysis once a year, because it’s such a big job, and you have to collate all the data and clean all the data and you’re going to make different types of decisions, then if you press a button, then you can see the current state of the data.
Denholm:
Exactly. And linking it to technology. I mean, when we work with product owners, and people are involved in developing new features for products, you know, what are the things really that help decide what that new feature should be?
Well, it’s the data that comes from the customer’s interaction with it, or, you know, the feedback that comes from a workshop, all that stuff is data. And it drives again, it’s just more frequent decision-making. So that might go up to a strategic level, yes, we need to release this product, or we need to release this new service or business line or something.
But importantly, it kind of rolls down to the rest of the firm who make decisions every day, big or small, how do you make data part of that as easy as possible.
And when you’ve got data that maybe you don’t trust, it’s not the right quality, the controls and processes around, it becomes really hard to automate that and you end up either locking it away, which means you’re not getting the value out of it, you’ve got an asset, and it’s worth X amount that is tucked away that you got to go and ask permission to use, or you’ve got a nice solution, basically, that opens that up in a trusted and governed way. That basically puts data as an asset in the hands of people across your business to make decisions because that’s their job.
Zoe:
Yeah, fantastic. And thinking about this kind of mechanisms for exploiting data or viewing data, tell me a bit about BDO’s data marketplace.
Denholm:
So the marketplace is really central to our strategy for context, you know, we’ve got eight and a half 1000 people, as I mentioned earlier, in the UK. And we know, as we’ve just discussed, that data is an asset for them. And I believe and I think a lot of the firm believe that actually, it helps them fulfil their roles and responsibilities more effectively.
It might be about efficiency, it might be high quality. And actually, many of our people across the firm, you know, they’re client-facing, or they go out to audited entities. And they’ve got analytics technologies or data visualisation technologies, they’ve got access to that already.
The bit that’s harder for them is, how do I get access to data that I know I can trust? Where is it? What does it mean? Is it of the right quality? So what we really wanted to do through the marketplace is how do we democratise data?
How do we put it in the hands of people across the firm, but do it in a trusted and governed way? How that works, it’s a one-stop shop, where you go into an internal portal, and you can search there’s weight first, that you would search for data. So it might be for us private equity clients or something like that, right. And it will bring up all of the relevant data products.
So immediately you’re solving that problem of, you know, I don’t have to know who might have access to that or in which system, it might be, I’m searching a business term for a business challenge or a business opportunity that I’ve got that I need for my role. And it’s bringing me back the relevant products as if I’m searching on Amazon or something, I can then request access.
So it’s controlled, it’s monitored, there might be some data products that we say, “Yeah, everybody should have these”, there might be some that we say actually only certain parts, and certain teams should have access to that. And again, we’ve now got control around that, because you can request access, you can understand the data.
So you see the products. And you can say, well, here it is, I can see maybe what it was used for before. I can see where it was collected in which system, that kind of lineage stuff, I can see what it was collected for it what I can and can’t use it for when you think about GDPR.
And so you go into it with eyes wide open in terms of what does that data actually mean, you can’t misconstrue stuff, you can see whether it’s fit for purpose. If I’m going in and I’m trying to build a regulatory report, I need that data to be 100% right?
If I’m doing some analysis, you know, taking through some concepts, it’s early stage, I might be able to accept less than that, because I’ve got the right context and caveats. And as part of the decision-making process, again, you go into that eyes wide open rather than finding out at the end that actually it’s 95% accurate, or something else.
And that’s not fit for purpose for you. And then the last bit is that actually, you just use the data there in the marketplace. So rather than having to download a CSV file or something, put it into an SFTP or a SharePoint folder or finding a way to send it to somebody who wants to do something in Power BI, you do it there, the data stays in our enterprise data analytics platform. And that takes off a big security concern, you know, you know, where the data is not moving around.
And, you know, it kind of lines up to one of our principles, which is that we fix data at source. One person tells us that something’s not right We fix that in the source system. And it flows through everybody else who wants to use that same data through the marketplace. So I sometimes talk about it as “governance by stealth” because governance, Scott, a terrible branding issue. But yeah, that’s the marketplace.
Well, it’s kind of themeless governance, isn’t it rather than some big cumbersome process. And I think, for me, that really eliminates this idea of like, using the data at source, I think one of my biggest concerns, if I’m ever dealing with like, a large body of data is introducing inconsistencies or introducing errors somehow, in how I’m transmitting it, like, missing off the second half of an Excel file, or you know, all of those kinds of things. So I can see that that’s really reinforcing, like trust in the data.
Zoe:
So could you tell us a bit more maybe about how effective data governance whether in this way or in other ways, builds the trust between data providers and consumers? And what are the kind of key elements that?
Denholm:
Yeah, so when I think about effective data governance, I think about is data quality improving, is our understanding of our data improving, is our use of our data increasing. That’s kind of what I think it’s about making it practical, it’s about exploiting and extracting that value from this asset. So that’s how I kind of think about it becomes more trustworthy, because the quality is improving, becomes easier to use, because I can understand what it actually means. And therefore we use it as part of more business process, more parts of our business model, and it drives efficiency or quality. So more returns and more returns, let’s be honest. Yeah, exactly.
So and I kind of talked about data governance, maybe having a bit of a branding issue before, but I think, you know, the word that I use most frequently with my data governance team is practical. How do we make this practical?
And it’s why I really liked the marketplace, it’s practical, you go there, you search, you find stuff, it solves the data, consumers issues and pain points. So, if I’ve got to go find a, and I don’t know whether I can trust it, and I don’t really know, it means take that away, and you solve their issue.
And you also introduce all the nice stuff around control and compliance with security policies, etc. So I think the practical bit is maybe the first one. The second bit, and I don’t know who said it, but there’s a piece that says basically, if you want to do data governance badly, you do data governance across all your data. If you can’t, you can’t govern absolutely everything.
And again, this kind of valuation piece with data tells you what your highest value data assets are. So you start there. And I think, again, making it practical for me is about work with the business partner with them, the users of the data, those consumers of the data, and solve what matters to them, I need to produce this report. So I can make these decisions.
This is an operational process that falls over because of poor data quality, solve that for them in the most MVP way possible, try it, learn, iterate. And over time, you’ll build up your massive policies and standards and processes that kind of complete your data governance framework.
So I think it’s almost taking that practical kind of business lead approach and applying your data governance knowledge to it. So I think those are the big two.
And then the third one, I think is, as with everything was trust is transparency, you know, they really need transparency, because if they’re gonna build a report, or they’re gonna build something, they’re gonna use data in some way. If you’re transparent with them with the issues that you know, the stuff that maybe you don’t know, then they can make a decision because ultimately, the consumer has to make a decision on whether it’s fit for purpose or not. They might say, okay, that means I can’t use it. But they do that knowingly. And they do that from without wasting effort on it. If you wait for them to find out at the end, because you want it to be part of the use case, you break that trust, and it might destroy the trust of more than that data set itself, it might destroy the trust of the whole platform, the whole marketplace. So being transparent and being like the data quality on this isn’t great. Don’t use it for regulatory report.
You can use it for some analysis. But beware upfront, you can decide whether it’s fit for purpose for your use case, versus this one, we trust. This is our authoritative view of client. And we stand by that we can trust that and go for it, use it for anything that you need to.
Zoe:
Yeah, so important. And looking more broadly, what would you say are the emerging trends in data usage and governance? And which of those do you think are going to have the biggest most significant impact on the industry?
Denholm:
Well Gen AI is the obvious one, isn’t it? So I won’t go too much into it, but it’s huge. And I think it’s very exciting. I think the bits that maybe hang off of that are the bits that I’m quite interested in.
I’m doing a lot of thinking around because I think Gen AI you know, it’s captured lots of people’s imaginations, it’s very practical. There are less barriers to using it. type some stuff into a seemingly a chatbot and it feeds some stuff back to you so execs to grads are getting really hands on with AI.
The stuff that I’m starting to think about around it as well. How do you make Gen AI I create a competitive advantage for you? So if everybody starts to use open AI or bar or some of the really big LLM ‘s that are very common, well, then it’s less of a competitive advantage, because it’s the same data that’s in the LLM.
And maybe it comes down to how you integrate it with your business model. But equally, I think it can be about how you build your own data solutions alongside that, and how you maybe build some smaller language models or something to augment what the big MLMs are doing.
So I think there’s something around how you make Gen AI be a competitive advantage over the longer term, there’s enthusiasm, I think, will seek out into kind of wider AI applications, like Gen AI is not the only AIP.
So AI has been around for a long time. And I think kind of one of my hopes is maybe that it encourages more kind of execs, who maybe to have a bit more confidence in taking a bit of a punt and a bit of a risk and putting money into AI and machine learning and advanced analytics capabilities. Because I can see, really practically, its potential through generative AI.
And then the third one for me is about, it’s going to turn its focus to data governance, it has to do the foundations remain the same for me. If you are going to be building your own small language models, you’re going to be building using other applications of AI.
You need good quality data. You know, if you’re talking about graph databases, you’re talking about neural networks, you really, really need to understand what does your data actually mean? And how do you connect it up in a way that you can trust a machine to go and learn from it. And I think that becomes really important.
And we’re gonna see regulation, obviously emerging in this space, it’s starting to happen. So I think getting good data governance done now means that you’re in a nice place to exploit AI more broadly in the future.
Zoe:
Yeah, exactly. And kind of coming back on the conversation we were just having about transparency and repeatability, it’s how are you going to be able to trust it? Unless you have that in place? Right? It’s kind of not really optional. All right. Just finally, quickly, are there any upcoming projects or initiatives that BDO that you’re particularly excited about in terms of data management?
Denholm:
The one that I’m excited about is the one that I mentioned at the beginning, you know, we’re partway through is the data evaluation piece. I’m really excited to see how the evaluation resonates with other people across BDO.
You know, we’ve got a lot of accountants in BDO to challenge the valuation. So that’ll be fun. I’m really excited about that. I’m excited about how it maybe changes the perception on how we manage data across the firm through its lifecycle.
And from a data perspective, I’m excited for how maybe it shifts the data office’s thinking because it’s so we’re going to get a valuation of realised value probable value from our data. But how do we shift that up? I think there’s huge potential invade.
So if you think, open AI as the example, multi billion pounds business model that’s just seemingly come out of nowhere overnight. Clearly, it hasn’t. But what’s the thing in the middle of that?
It’s not the stuff that’s on the balance sheet. It’s not the number of employees, it’s not the buildings that they’ve gone. It’s not necessarily the technology, that’s part of it. The middle of it, is data.
And I think if we start to realise that a BDO leaders in the mid-market, beacon network of people who are collecting data every single day, interacting with clients getting leading expertise in an information economy, I think can be super exciting. So yeah, I’m excited about the evaluation.
Zoe:
I’m now excited about Data Valuation of Well, thank you so much, Denholm, for joining us to share your insights and also to help us shine a light for others.
Denholm:
Pleasure. Thanks for having me.
