What If Data Was a Product You Could Sell? An Interview with Eric Walk. 

Eric (00:00):

Okay, so we've made the enterprise scale decision. We're going to invest in this, we're going to make the operating model changes. We're going to assign that accountability and responsibility for product ownership, for data. out to the business or, or whoever. But usually, the businesspeople who create that data, own that data. We’ve invested in the tools, we've stood them up, and we've implemented them. We've built out policies and procedures to govern everything. That's all great. If you build it and they don't come, you've got a problem.

Jim (00:32):

Welcome to What If? So What? the podcast where we ask what's possible with digital and figure out how to make it real in your business. I'm Jim Hertzfeld.

Kim (00:42):

And I'm Kim Czopek.

Jim (00:43):

And today we'll ask, ‘what if’, ‘so what?’ And most importantly, ‘now what.’ 

Jim (00:48):

You've probably heard stats like this before, but at this point in time, 2.5 quintillion bites of data are generated every day. That's a billion, billion. It's the natural consequence of years of hyper-connected technology that's generating data from everywhere, from social media, e-commerce, IOT devices. By the way, I looked that up. That's 14 billion as of today. And of course, everyone's favorite source of data, large enterprise corporate platforms. Aside from the frankly unsurprising growth in data, what's a little bit different is it's no longer exclusively owned by its creators. Individuals and organizations are generating and relying on data across a bigger variety of platforms and sources. It's everywhere. People are struggling with how to get it, who to call, and if they can trust it. If you go way back to the beginning, early IT, or literally data processing, you would literally print reports, if you know what I mean by a green bar, you just dated yourself. It stayed that way for decades. But this massive growth in data, it's an even bigger problem. Of course, a bigger opportunity that we want to talk about today. But, guess what, it's no longer just an IT problem. So, our ‘What If?’ today is, given this explosion in data is what if you could get the data you need to fundamentally do your job differently? On an individual basis, Kim, big promises a lot of excitement. So what?

Kim (02:13):

Yeah, I'm skeptical. We've talked a lot about data with our guests over our episodes and we always seem to arrive at needing insights over data and I don't think anyone will argue that. And 20 years ago, 15 years ago, 10 years ago, even today, there are aspects of my job where I can pull my own data and I don't get the insights that I need to do my job differently, let alone change the way I work for the better. My ‘So What?’ is: is data management really evolving or are we just seeing another set of buzzwords? But the good news is with us today to help us figure this out, is Eric Walk, Data Strategist at Perficient. Eric is going to help us figure out what is new in the data space. And in particular, Eric, I keep hearing this phrase data mesh as the new promise of data democratization. Can this really help us change the way we work?

Eric (03:09):

Thanks Kim. Data mesh is an interesting idea. It's something that you hear a lot about, out in the world, and it's often quite misunderstood. Ultimately, it's really technology agnostic. It's not even really about the technology, it's about the way we work. It's about the way we think about data and it's about the culture of our organizations and the way we interact with data, as we're thinking about the principles of data mesh. Really, we're shifting to this product-oriented concept, where data's no longer an asset, data is no longer a burden. Data's not something that sits in a warehouse. It's a product. We sell it, whether we're selling it to the outside world in some organizations. But, in most situations, your customer, your consumer for your data products is all internal. And with that idea, right? We start to think about, okay, I have my data, I produce it, it's a product that I sell, that I provide to my colleagues and other departments and other teams, they do the same. And in that way, we can finally start to get in a controlled way, access to trusted data that are known to be high quality, that's known to be accurate in a consistent way so that as we're starting to build insights and think about the questions we want to ask and use our data to answer, we have the ability to say, yes, I know I'm getting the right data. I know I'm getting it on time. I understand its characteristics, and I trust that it's been reviewed and validated by the people who know it best. When we start to think about all of this and the philosophy of data democratization here it gets sort of high concept, right? We often start to lose track of what we're really doing. So, bringing it back in, the question we ask is, why do I need the data? What am I doing with it? So, let's say, I want to answer a question about my sales. I want to know how many customers have purchased a particular product. Did they buy it on financing? When is their payment due? A whole bunch of different aspects of that purchase that are interesting to me. Well, I might be in the sales department, and I only have control over whether that person came in or the data that shows whether that person came in and made a purchase. Right. I don't know if they used financing. I don't know when their payments are due. I don't know anything about what happened after the point where they made that purchase. So now, I might need to go out to all sorts of different parts of my business and ask them, hey, do you have this? How do we track that? How do I connect them together? If we start to build on these data mesh principles, we start to build out a marketplace. We start to build out a scenario where I can, in an e-commerce-like way, go into the portal and say, hey, I need this data about who bought this on financing, when their payments are due, et cetera, so that I can make some best guesses about other characteristics of these people, what else they might want to buy. Great! If I have a marketplace where I know that data's managed, it's trusted, it's accurate, and I can request access to it, all of a sudden, I can accelerate my use of data beyond what's available inside my department to what's available within other departments in a controlled and reliable way to start building those more complicated insights with data that's outside my realm of control.

Jim (06:33):

Eric, is that really what you're getting at, out with democratization? Do you mind just kind of building on that a little bit? I think that is a high-concept word.

Eric (06:40):

Yeah, when we talk about democratization, what we really mean is that we want every person in a business, no matter how senior or junior, to be able to feel confident working with data and using data to help them make better decisions. You know, we talk a lot about insights, but why do we need insights? Right? Ultimately, it's about driving data decision-making. You know, again, thinking about my sales department. I want to know which clients I should prioritize. I want to know who's most likely to make a purchase. I need to make decisions about how I spend my time, where I put my effort, about where I spend my money and using upstream data, high-quality data to help us drive that decision-making is really the key. But our businesspeople can't do that if they don't individually at that individual level as you said in the intro, have access to that data and the confidence to use it, with trust, with really strong literacy and the ability to do it independently.

Jim (07:38):

So, it's not about voting.

Eric (07:39):

No, it's not about that.

Jim (07:41):

Okay, just want to clear that up.

Eric (07:42):

You know, the words I always fall back on right, are availability, discoverability, and trustability, if that data's not nutritionally available to me, if it's not in a format I can consume with tools that I know how to use, I'm still not able to get value at it. I'm not able to drive insights. I'm not able to use it for that decision-making. If it's not discoverable if I don't know what it is if I don't know where it's coming from if I don't have that metadata, the data about my data. I can't figure out what I need to use. I can't find what I'm looking for. I can't get started. The biggest problem that I hear people talking about today is that they don't trust their data. The quality is poor. They know they have the data. I think about a person I met last year, who is an industrial engineer in a packaging plant, and he can't figure out if it's worth it to automate the packaging of a particular kind of product, it's done by hand today. He knows he has that worker efficiency data somewhere. Someone must have it right, but he doesn't know how to get to it. He doesn't know if it's high quality. He doesn't have any information about it, so he can't find it. It's not available to him. He can't consume it. And he doesn't necessarily trust it's right, and so he can't go make the business case to his boss to invest a million dollars to automate that process. So, when you get down to it, right? In this case, literally nuts and bolts the question is, are we able to use day to make decisions in a day-to-day way? Even if you're the type of person whose job involves a hard hat and, and steel-toed boots. How are you able to use data to make better decisions about what you're doing and how you're prioritizing work and how you can make your job more efficient, more effective, more profitable for the business? It's not easy, right? We get down to how at a certain point, and this is the place where the challenges that Kim was talking about really come to bear. This requires a culture shift. It requires organizations to think differently, to work differently, and really go and make a strong enterprise-scale investment in governance, in ownership, in that product-oriented thinking. We talk about product-oriented thinking, right? It means we need to assign product ownership to someone within the business who's going to drive the shape and size and nature of the features of that data product. We need to assign a product development team who's going to do that work to make the product meet the requirements. We need a set of controls to make sure that the product is abiding by regulations and working the way we want it to, providing the features to the consumers in a way that's safe, secure, appropriately private, et cetera. We need to make investments in tools to help us automate that, whether it's a data marketplace platform a data catalog or other kinds of tooling. Once you're there, okay. So, we've made the enterprise scale decision. We're going to invest in this, we're going to make the operating model changes. We're going to assign that accountability and responsibility for product ownership for data out to the business or whoever, usually, the businesspeople who create that data, own that data. We've invested in the tools, we've stood them up, and we've implemented them. We've built out policies and procedures to govern everything. That's all great. If you build it and they don't come, you've got a problem. Adoption is key, working with our consumers of data, our producers of data, and our end users, and really treating them as the customer of a product is a critical piece of the puzzle. I said in the beginning, we want to create that e-commerce-like experience where someone can go in and add data to their shopping cart. Right? And use it to start to drive insights with confidence that it's accurate, trustable, and that they're going to be able to use it within tools and environments that they understand and they're comfortable with. In order to do that, we have to really market it. We have to use our change management muscles, our organizational change muscles, and our behavioral science muscles, to make sure that we're creating journeys for those users that will bring them along and encourage them to do the work. This isn't something that an IT team or an individual business unit can do in a vacuum. It really requires that large commitment from an organization at all levels in order for it to really be effective. The other piece of this is we never talk about implementing a data mesh. We don't build a data mesh. We don't make a data mesh. It's a set of principles, a set of ideas, right? We have to tailor the solution, operating model, and organization model to an organization, to an enterprise in order to make sure that we get that adoption, that we get that engagement from our employees to really drive the success of this kind of program in the long run.

Jim (12:44):

Eric, what I'm hearing too is, it’s not just the adoption, but awareness. I think goes back to the “what if?”, What could I do? Maybe there's a lot of education on not just what's out there, but a lot of new skills for people as in how do they utilize data?

Eric (13:02):

Exactly. It's new tools, it's new ways of thinking. It’s making sure that people feel confident when they know how to use statistics and do so in a way that they're comfortable. They're not lying, they’re not using those statistics to show something that isn't true. They don't just need to trust the source data, they need to trust the process, the tools, the algorithms that they're putting in place, that they're using to drive those insights and make those decisions.

Jim (13:29):

Hey Eric. Thanks for the insight. I think it's definitely a different way of thinking. Just to kind of close this out, we're hearing from a lot of companies every day that they want to become technology companies. We hear that all the time, how we are using digital to go from this legacy of a hundred-year-old company to a technology company. Maybe the new message is whether they know it or not, they're becoming data companies, they're generating data and really kind of becoming everyone's problem. If you're not sure about that, just take the stock of the data that you're creating or consuming, whether you're sending a message uploading a file, clicking a button, or asking for a report. It's an ongoing, continued generation of data that may have value we haven't even conceived yet. I think it’s something for folks to think about. I was keeping track of the data that we've created today. We're about 15-16 minutes into this. At our capture rate, I’m going to go ahead and put us down for 158 megabytes of data on this episode, we hope you tune in for more data in the next episode.

Jay (14:36):

You've been listening to What If? So What? The Digital Strategy Podcast. From Perficient with Jim Hertzfeld and Kim Czopek. We want to thank our proficient colleagues, JD Norman and Rick Bauer. For our music today, subscribe to the podcast and don't miss a single episode. You can find this season along with show notes at perficient.com.

Thanks for listening.