Champions of Data and AI

Episode 8: The Critical Job of Building a Data Culture

July 13, 2021 Databricks Season 2 Episode 1
Champions of Data and AI
Episode 8: The Critical Job of Building a Data Culture
Chapters
Champions of Data and AI
Episode 8: The Critical Job of Building a Data Culture
Jul 13, 2021 Season 2 Episode 1
Databricks

Culture can make or break organizations. It also plays a vital role in an organization’s ability to become data-driven. In this episode, sit down with not just one but five data leaders in a panel discussion on building data cultures. 

Show Notes Transcript

Culture can make or break organizations. It also plays a vital role in an organization’s ability to become data-driven. In this episode, sit down with not just one but five data leaders in a panel discussion on building data cultures. 

Speaker 1 (00:00):

Welcome to champions of data and AI brought to you by Databricks. In each episode, we salute champions of data and AI, the change agents who are shaking up the status quo, these Mavericks are rethinking how data and AI can enhance the human experience. We'll dive into their challenges and celebrate their successes all while getting to know these leaders a little more personally,

Chris D’Agostino (00:33):

Welcome to season two episode, one of the champions of data and AI series. I'm Chris, D'Agostino your host for today's episode. Culture is what makes or breaks organizations. It also plays a vital role in an organization's ability to become data-driven. In this episode, I sit down with not just one but five data leaders in a panel discussion on what it takes to build a data-driven culture. Everyone, thanks for joining the panel today. And we've got a list of questions around strategy and how organizations are actually moving forward with data and AI and how it's changing their business. So let's dive right in. So the first question that we have is, you know, I'll call out individual panelists to give the first response and then some follow-up responses, but we want to talk a bit about setting the vision and the goals for your strategy. And so generally speaking, Yao I'll turn to you first is what are the most critical business objectives for your company's enterprise data strategy over the next two years? And then also what are some of the driving forces behind creating that strategy?

Yao Morin, JLL (01:45):

Yeah, to put it simply, JLL is a commercial real estate company and our main goal for our enterprise data strategy for the coming years is to really, to understand how to unlock the data and intelligence to deliver robust and real-time insights for commercial real estate. Um, probably not many of you knows that world commercial real estate is the largest global investor class and the, we are actually stewing a very early stage for tech adoption and, you know, it's actually the second largest expense for business. However, we are so lacking the data and the insight to help companies to make important informed business decisions on commercial real estate. And I believe that, you know, data can really make an impact and, and have that real world, uh, important impact on, on everyone's life, especially, uh, in this COVID 19 world. Uh, if we will be able to deliver that data and insight, we can actually help, um, companies and employees or employers to one is to reduce the environmental impact and also facilitate a really safer ways of working in this COVID world. So very, uh, you know, very big ambition, very big innovation, and we have a very long way to go.

Chris D’Agostino (03:17):

Great. Thanks. And Patrick next, and then Randy, if you don't mind.

Patrick Baginski, McDonald’s (03:23):

Yeah. Uh, thank you for having me. So at McDonald's I think the way we think about it, it's, um, the best way to, to address your data strategy is to make sure that you truly derive value from data. And there is a, there's two critical components to driving value from data to us. And one is just the closeness to the user, the availability of it, and then ultimately also data governance. Right? So when I say user, I don't just mean the data scientist. I actually also mean, you know, the business analysts, the marketers, um, you know, field operators, so as close as possible to the user and you need to create that trust in the data too. Right. So for us, it's really all about, you know, data governance, creating broader availability and making sure that users can meaningfully interact with the data too.

Chris D’Agostino (04:20):

Yeah. And before, before we get to Randy, I want to follow up with you Patrick and yell a bit, you know, your industries have been particularly hard hit with COVID right. Restaurants have shut down, uh, for, in, in restaurant dining. Many of them, you know, had to reconfigure their restaurant offering to have takeaway. I mean, obviously McDonald's has that built into built into their stores for, for many of them, but for you, Randy, and I'm sorry for you, Patrick, and for you. Yeah. Well, with the change in how people occupy office space now, and the social distancing requirements and the ventilation requirements, how have you seen the data that you're collecting help drive, you know, the revenue streams that, uh, that your businesses rely on?

Patrick Baginski, McDonald’s (05:07):

Yeah, so I actually think that, um, from a, from a data perspective, uh, COVID accelerated the journey for us, but it didn't necessarily transform it. I think we're, we're past the stage of transformation and we're at the stage of, you know, truly accelerating what we can do with the data, what everybody can do with it. And so, uh, we naturally focus on, you know, digital delivery and drive through and continue to, you know, focus on these data sources as well as they're driving the biggest, you know, ability to drive impact and value for the business too. Right. And so, uh, the, I would say that focus has become bigger, uh, over the COVID pandemic, but it also has become more important than more prevalent in people's minds. Right. And so you see that, uh, now users in the organization is starting to think much more holistically around, okay, how can we valuable value for our customers and our restaurants use this data now, right?

Chris D’Agostino (06:13):

Yeah. Quick question there, Patrick. So it seems like if the encounter in restaurant experience changed and people aren't walking up to the counter to order their meals, or walking up to one of the digital kiosks to do that there. Now, if they have to stay outside of the restaurant, they're now using the mobile app. I'm just trying to think through like the delivery flow of food, uh, given the cues that would, would be created inside the, you know, the drive-through lanes have limited number of drive-through lanes. Of course, you're not gonna build new lanes during COVID, but you're probably taking on more, more mobile orders to replace that in store counter experience or kiosk experience.

Patrick Baginski, McDonald’s (06:55):

That's that's correct from a data perspective. And, but there is a little bit of a, of a balanced by market, right? So every market has a little bit of a different adoption rate and a little bit of a, of a different distribution by channel. But you're right in the sense of that, we're obviously trying to further integrate all the all those different data sources from all the different channels, and also be able to, uh, find new use cases, find new analytics, new data science work, and how we valuably can, can use this data.

Chris D’Agostino (07:28):

Great. And Yao circle, back to you, you know, how has the change in how real estate is used commercial real estate is used given COVID? How has that affected how you analyze the data?

Yao Morin, JLL (07:39):

Yeah, so, um, it's, it's a really, it's a challenge, but it's also an opportunity to, to really accelerate how we use data in commercial real estate. A lot of people think that commercial real estate is really slow moving. Uh, you ran a place, you you're there for a few years, but now because of COVID, um, everyone is rethinking, how do we, um, how do we change our relationship with the space that we are in? Um, a lot of companies start thinking about, you know, can we have smaller spaces everywhere and to provide the flexibility for our employees and how do we reconfigure our office into having more of, um, a coworking space where people can shop in whenever they want to be meeting the coworkers. And then, um, the other, the other more office desk work, they will be, um, you know, doing at home.

Yao Morin, JLL (08:36):

So our data and analysis is really helping our clients to understand their space and then to plan, um, better in terms of providing the safe and flexible work environment for our, uh, for the employees. And I think that more than ever, they realize just how it is to have those data at your fingertips so they can respond really fast whenever that hit. I think that COVID really is a wake up call, um, to people on how important it is to have those data and analysis. And, you know, it's, it's good for us as data practitioners. We don't need to sell how important data is anymore because they would know it and feel it.

Chris D’Agostino (09:19):

Yeah. Yeah. And Randy, you've written a book on this subject, so can you give us some of your insights on what these critical business objectives are for companies that you're studying?

Randy Bean, NewVantage Partners (09:32):

Uh, thank you, Chris. I'm delighted to be here. Uh, first of all, I echo the comments of both Yao and Patrick, in terms of the importance of business value. That's really what the investment in data is about. Uh, the perspective I bring is working with a range of organizations and saying what a range of organizations are doing and some of the challenges and opportunities, and to echo your statement. I think data has become more critical now, more than ever. And a lot of this has been elevated by the existence of COVID and the greater public awareness, for example, of the COVID dashboards. But I want to share some data with you, some statistics from a survey that we conduct each year of leading fortune 1000 companies, which presents some of the challenges that organizations still face. So we ask these questions. Yes or no. Have you created a data-driven organization? Only 24% responded Yes. Have you forced a data culture? Only 24.4% stated yes. Have you achieved transitional business outcomes? Only 29.2% reported yes. Are you managing data as a business asset? Only 39.3% replied yes. So I could go on with these statistics, but they point to some of the challenges that fortune 1000 companies continue to face in terms of being able to drive business value and leverage data as an asset.

Chris D’Agostino (10:56):

Yeah, Randy, thanks for that. I think this is a good time to sort of layer in Sherman and his perspective. You know, we set these goals and these organizations, they want to leverage more data. They want to leverage data between business units. MasterCard of course, has done a great job with their data strategy. And so want to talk to Sherman and get his point of view on what defines a data culture. And what impact does that have on how people within your organization spend their day at work?

Sherman Cooper, MasterCard, (11:26):

That's a fantastic question. The light, it could be here. And thank you guys for this amazing forum. You know, when I think about a data culture, you know, that's a really peculiar notion in my mind, right? Because data one is contextual, but also cultures are unique. So when I think about a culture, an organization's culture, I really think about the rules, norms, and values that sort of underpin the strategy, the business, and what is the business trying to, and for whom. And so sort of hearkening back to some of the earlier comments around value. Uh, when I think about a data culture for me, it's really understanding how do the data that we collect and how we utilize those data. How does that ultimately help to strengthen the culture, those rules, norms, and values that we're talking about? So, so the data have to deliver, uh, right, some, some, some value proposition around maintaining those norms and sort of moving toward the, uh, the in strategy, right for the organization, you know, at MasterCard, we're a culture of decency.

Sherman Cooper, MasterCard, (12:28):

Uh, we're really focused around using our payment network, uh, to power the transactions that make lives better, uh, across the world where we do business in over 210 countries in territory. So our data philosophy is, is both defensive and authentic. You know, we use our data, uh, to make people's lives better, to develop new products, solutions, and services that are human centered, and that, that offer new efficiencies to our customers and our partners, uh, and that ultimately, you know, improve, improve commerce and, and, and do that in a secure and responsible way, uh, for, uh, for everyone. So, uh, when I think about how does this impact our day-to-day life at MasterCard? You know, our data officer is a brilliant, uh, luminary. Her name is Joanne Stonier, and she, she really leans on us a lot, you know, about making sure that, uh, as the data office that we're close to the business.

Sherman Cooper, MasterCard, (13:16):

And so we're actually embedded into the business at MasterCard in multiple ways. Our strategy function really is tied to having an expert, right, who is sort of aligned with, with each area of our business to ensure that as new product solutions and services are being developed, that we're at the table thinking about the implications of, of the business strategy and the new product on our data strategy. So we sort of really walk hand in hand. So our culture at MasterCard existed, uh, right. Uh, obviously before we thought of ourselves as a data company, even though we always sort of been a data company in many ways, but, uh, I think now we're leaning into making sure that the value proposition around the data that we collect, the data that we use, uh, that it's been in a responsible way, that that's really tied into our culture of decency and that's really improving lives across the globe. So, great question.

Chris D’Agostino (14:07):

Yeah. Sherman want to follow up with you here. Wanted to find out, you know, in, in a lot of organizations and certainly ones that I supported in the past, um, in, in my career, there would be talk within the organization about a business objective, and then oftentimes times one or more projects that would be built to support that business objectives. And they often, you know, they usually had some type of application being built and we would refer to the application or the platform being built, you know, and we would give it some fancy name and everybody would say, you know, did you, you know, are you using system X, Y, Z, or whatever I'm curious about if you've seen a shift where people are thinking less about the platform of the application that they interact with and about the actual data assets and what data assets they need, what data assets that platform works with, you know, does the language within an organization, the size of MasterCard start to evolve to thinking about data as an asset rather than the systems that are being built to serve up the data?

Sherman Cooper, MasterCard, (15:09):

That's a great point. So we've actually noticed that. So, uh, you know, I will tell you from the very highest levels of our organization, uh, right, there's no lot of conversation right. About, you know, data as an enabler. And so when we think about data and MasterCard, we're really talking about data as an enabler of our business strategy, right. Rather than platform X or platform Y now granted, right? Because we live in a world of GDPR and CCPA and to be CPRA and all of the other respective, uh, privacy and data protection regulations around the globe, we do have certain platforms that allow the transparency required, uh, not just at the legal standard, but above that, right, to engage, uh, you know, consumers around the globe, but individuals to allow them access to the information and control over that information with respect to MasterCard's use of data.

Sherman Cooper, MasterCard, (15:58):

Uh, but you know, beyond those platforms, right? Well, we have several at MasterCard. The conversation is less about the platform and more about business enablement. And I think that shift in language is really, really important because it shows to the point of the earlier panelists, that we're now leaning into a bigger value proposition, right? Where the data are closer to the business, right? Folks we're using the data every day and that the data are really enabling better business decisions, really helping to bring the strategy to life because MasterCard is a global dynamic organization. And certainly right. We've seen an increase in the use of path and go right. Tap to pay, right. A lot less swipe were happening during COVID, right. We've obviously seen, you know, a shift to online purchasing as opposed to in store purchases. Right? So, but when we're thinking about how we can develop better products, solutions, and services, uh, to sort of enable, uh, this, this, this transition that we've seen and to not just enable it, but to help it thrive and think about how we build better products into the future and really build the next generation of payment products and solutions, uh, really it's about enablement.

Sherman Cooper, MasterCard, (17:08):

And that shift in language has been prominent at MasterCard from the very highest level. And you really hear business leaders at all levels of the organization talking about the data that they need to enable this product or to enable this initiative. So it's been a refreshing shift, uh, but you're spot on there.

Chris D’Agostino (17:25):

Great. Yao, what's, what's your point of view here in terms of defining that data culture and, and how people within your organization come to work and show up at work differently. Now,

Yao Morin, JLL (17:37):

I think Sherman covers a lot of the ground. I felt like, wow, it's true. I, I swear I didn't actually call it. He knew it was show, but in about answer, but definitely it's not about the technology anymore. It's really about how do we use data in our day to day, um, in our day-to-day business and then making that data driven decision. Right. I think that one of the key thing I think just to add on to Sherman's answer is to really think about data is not just for a centralized data team or the data practitioner's job is really everyone's job to think about data and then take a data first approach and to think about, Hey, how can I leverage data to actually do my business better, smarter and more efficiently? Um, but yeah, well, Sarah Sherman,

Chris D’Agostino (18:33):

So Don, I wanna weave you in here. I'm sorry. We haven't heard from you yet, but, I wanted to get your point of view, you know, you likely are in an organization, that's got different business units that have a particular view of a given customer. How do you see the culture and, you know, really sort of driving that culture to get everybody across those business units to share data more willingly, um, maybe break down some of those silos so that you can look for opportunities to cross sell products within your space?

Don Vu, Northwestern Mutual (19:05):

Yeah, no, I mean, I think Peter Drucker has that quote culture eats strategy for breakfast. And I love this, uh, survey that you noted the New Vantage Partners, one that talks about culture as the main impediment for actually executing and delivering on AI and ML initiatives. I think that was actually the number one, answer five years in a row. And it's something that I've actually shared with my own CEO. Um, it's really the main impediment for actually us delivering value. And really, I think one thing maybe I'll just unpack before I answer your question was around. Why, why is that a challenge? Um, and why has that been a challenge for many, many years? And I think it comes down to a lack of understanding and a lack of knowing what the art of the possible is for data. And, um, oftentimes that means data literacy and data literacy for us is partnering with our business partners and really trying to empower them and everyone at Northwestern Mutual to understand how you can solve problems using data.

Don Vu, Northwestern Mutual (19:56):

Uh, and that's been really key for us really, uh, having that partnership and collaboration such that we can together drive forward business outcomes. And I've seen that just time and time again and over my 20 year career with, uh, over 13 years at Major League Baseball and then about a year and a half at WeWork, uh, before coming to Northwestern Mutual and the problems aren't the same in data, but they do rhyme. Um, and so oftentimes what I've seen when in talking to our business partners, if you can actually give an analogy, something that they can really latch onto and contextualize for their own business problem, it can really help with that unlock. Um, but going back to your, the question you just, uh, phrase was around, how do you get people siloed in the organization to kind of work together and have more of an enterprise point of view on how to unlock data? Um, and I found ways found that, uh, anchoring and, and what's best for the organization as a whole what's for our customers. Um, and really, uh, reminding folks that our customers expect the holistic experience across all these silos has really, really been useful for us moving the ball forward with that. So that's been good for us. Did you,

Chris D’Agostino (21:00):

Don, did you do anything, um, you know, in terms of bringing stakeholders from different lines of business together who might not normally interact in their day-to-day jobs and trying to figure out, you know, I I've been in an organization, one of the top 10 banks where each of the lines of business was basically running its own PNL, had its own sort of objectives. And so from an enterprise data sharing strategy, trying to get them to talk together and figure out how exchanging data and of course being compliant with all the rules and regulations, but exchanging data could help drive each line of business to some new Heights. So I'm wondering if there was any particular approach that you used. We use things that have happened to one called hot houses where we'd bring people in and sort of phones down laptops down for the day. And it would be really kind of a very dynamic environment to exchange ideas and build relationships that maybe didn't exist really, uh, up until the, the data strategy movement. Yeah,

Don Vu, Northwestern Mutual (22:02):

We actually, and one of the first things I did when I arrived at Northwestern mutual was create a data strategy steering committee, that's fan all of our business lines. So whether it's marketing or our financial representatives, um, and other areas of the business with alongside folks from technology, as well as for our enterprise architecture group, legal compliance privacy, I think there's roughly 15 senior executives that are involved in that steering committee. We meet on a monthly basis. They review a moat roadmap. We talk about the various use cases that we unlock, and we've really found that to be a great forum within which we could just really pressure test our final roadmap. Um, and that's been great. And then it's also been a great forum to really underscore the value of having a unified data platform within which we create governance, access controls that although we have more work on the front end and ultimately creates a paved road for everyone else to actually deliver on business value. Um, so people really understand, uh, just the benefit down the line by doing a lot of this work together on the front end.

Chris D’Agostino (23:03):

Cool. Well, Randy, let's shift gears here a little bit and get you back into the conversation. You had a, uh, survey that you released in 2021 called big data and AI executive survey from new vantage partners where you're the CEO and you have a quote here and I'll read it. We're 92% of the respondents said building a data culture as a process, leading companies continue to identify culture people, process, organization, change management, as the biggest impediment to becoming a data-driven organization. So can you help us understand, uh, what those impediments are like? You know, what, what are sort of at the core of why the, that culture and, uh, identifying it is, is the impediment here?

Randy Bean, NewVantage Partners (23:48):

Absolutely. And I, I wouldn't know that, uh, having worked with both MasterCard and Capital One, those are two organizations that I commend as being among the most data-driven that I've had the opportunity to say. So to your point, uh, culture is really the challenge. Most organizations, it's not technology and that by so many organizations I walk into, I meet with the data teams and they talk all about the very robust capabilities they've created. And I meet with the technology folks and they say the same thing. They'll show me the capabilities they've created. But then I meet with the line of business CEOs, the executives responsible for the ultimate business outcomes. And often what they share with me is a lack of confidence in the data. Uh, they don't feel that they're getting that they get the data that they need in the timeliness that they needed, or the items that they need to make their key decisions. So there continues to be a series of challenges in terms of getting the business constituents who have ultimate responsibility for revenue, generation, market growth, comfortable, and getting the data, getting them that they do, they need in the time that they needed and giving them the confidence. So often I encourage organizations to start small focus on one or two business questions, establish some trust, build some credibility, create some momentum. And that's a path that you can begin to show that there's business value that you can deliver to your primary business owners and sponsors.

Chris D’Agostino (25:20):

Yeah, that's great. Well, you heard it from Ali and Arsalan in the prior session where they were talking about the importance of data governance and things like that, uh, in terms of, especially as it leads into machine learning and the ability to, you know, train models with data that you can rely on and that you can trust. And I think Arsalan in his slides demonstrated like, you know, just showed like how small the ML algorithm portion is for the overall amount of work that needs to be done to do machine learning well, and that most of it is data wrangling and all the data prep work and the governance side of it. So let's talk a little bit about the architecture side here. You know, Ali mentioned the Lakehouse, so it's this notion of bringing together the best of both worlds from an enterprise data warehouse, which oftentimes gives you very, very robust data management and good traceability of the data, uh, but maybe a more limited approach to the range of data assets that you can work with.

Chris D’Agostino (26:18):

And what types of questions you might be able to ask of the data, to the data lake, where you've got a broader set of data types that you can work with, but it's maybe less Matt well-managed. And so the Lakehouse concept is we're going to take the best of both worlds. We're going to combine them. And I know Patrick, uh, you've, you've worked with kind of the Lakehouse architecture and there were a proponent of it. Can you talk to us a bit about how those ideas from EDWs versus generic, uh, data lakes are combined and in you're able to leverage the data in a, in a better way?

Patrick Baginski, McDonald’s (26:52):

Yeah. Happy to, and I do want to say it actually ties very neatly into what Randy said before. Right? So I, I, I couldn't agree more. I think that data governance is critical to being able to deliver business value. And when we talk about impediments, I, I often think about this as impediments can be real, such as, you know, the data is just not there, or it hasn't been ingested, right. Or it's just not a real time or it's too asynchronous. So those are real impediments that that can be solved technically. But I think the bigger impediment that that was referred to is really this, this cultural change. Right. And it's more impediment in the mind in being able to create that trust, to create the, the ability for, for these business lines to consume data and actually derive business value from it. Right. So the way we think about it is that it's not just about, again, putting those technologies in place, having the data available, and then, you know, having data science teams, building good capabilities, as you say from it, uh, both data science and machine learning operations teams, but we actually think it goes beyond that.

Patrick Baginski, McDonald’s (28:02):

So we, we have an area that we call enablement. It's really data analytics enablement, and we're entirely focused on, uh, essentially training and working with the businesses and the business lines together to, to gain that trust, gain, that ability, gain that confidence in the data to be part of the process of data governance, be part of the process of learning to use the capabilities and applying the capabilities and ultimately, uh, together with them be part of the business value. Right? And so where the Lakehouse for me comes in in that perspective is that it helps us decrease this timeline to value that I often speak about, right. I manage, I said before, it's, it's about acceleration, but what it really means for me is that how quickly can you be from a business question or a business problem, or just an idea how quickly can you get to a level where the business line can actually use the data and use the capabilities to deliver value, to actually get to the, to the monetary or the time value on that.

Patrick Baginski, McDonald’s (29:08):

And so the Lakehouse, um, that in, in some of our projects and use cases, we started using that. And I gotta be honest, it's, it's really very early on. We've just started with it a couple of months ago. It helps us shorten that timeline, and it helps us bring that connection to the business, right? Because now it's no longer just a data lake or a data layer where the data scientists are fairly familiar and comfortable in creating new data assets through transformation and through certain, uh, exercises that they do in order to build good models for example. But now that same data is actually also there and trusted and consumable, let's say in a SQL manner by the BI teams, by the business teams in a very simple interface, right. And interestingly, what we have seen on top of that is that when we do use a Lakehouse architecture, and when we do actually rely on some of the SQL workloads that we're actually seeing very high gain on efficiency on, on, let's say some simpler data science, workload workloads and their deployment, and it just makes it quicker, right? It helps you bridge the gap between the users and the business that typically consume the warehouse and the users in the business that typically consume the data lake. And now the two are the same thing, right? So it becomes much easier to create that trust. It becomes much easier to create that governance and it makes it essentially shorter to get to business value from projects and use cases.

Chris D’Agostino (30:41):

Yeah. I can see, uh, Patrick there in particular with an organization like McDonald's where time to market is so critical, right. You've got lots of competition out there. Um, and you're, you're obviously trying to have repeat customers and drawing new customers, Don, from your perspective, right. Maybe a slower moving, uh, vertical, if you will, in terms of churn customer churn and things like that, help help us understand how the Lakehouse might impact the ability to adhere to CCPA or GDPR or any of the other governance, uh, requirements. Yeah,

Randy Bean, NewVantage Partners (31:17):

Well, I will say, like from a churn perspective, we certainly are relatively slower and, and we have less to worry about from that perspective. I think we retain 97% of our customers year over year from a policy perspective, but that being said, like our desire and need, and the urgency to actually deliver very quickly is what drew us to a Lakehouse architecture and the notion of a unified analytics, data analytics platform that serve as both descriptive and predictive analytics. Ultimately we need to be in, we want to be an AI and ML driven company. We're really on the innovating in the way that we do underwriting and leveraging AI and ML in a way that the industry never has before. And that's really a response to COVID. And that's just really in a response to just the overall customer experience needs that we have. Um, but really just leaning into the Lakehouse architecture.

Randy Bean, NewVantage Partners (32:07):

And again, this is unified data analytics platform has allowed for us to better manage the fact that we're aggregating data. That's regulated by HIPAA. We're aggregating the data that's regulated by FINRA and the, we need to have extremely fine grain access controls to ensure that even though it's all in one place and our client data is enabled to provide a holistic view across all the silos that we talked about before it's being done in a compliant, uh, legal and privacy compliant way, and again, creating updated road for delivery. So we've actually found like platforms to be an accelerant for business delivery. Again, it's a little bit more work on the front end, but after that, our ability to actually deliver value for the business and then subsequently, um, has really been accelerated.

Chris D’Agostino (32:54):

That's great. Yeah. I mean, one of the things that I experienced was when you had data sets in different data platforms, if you didn't have that unified platform, you often times were seeing the data curation steps in the data management, taking place in these separate environments and oftentimes repeating the same work. So, uh, teams were reinventing the wheel. They weren't being able to leverage the upstream work of another team. Uh, that's already doing some work to make the data more consumable. So it sounds like as you move more towards this paradigm of a lake house, uh, if you do it well, and if you organize the teams around that, and everybody understands kind of the inputs and the outputs and their producer, consumer responsibilities, as far as data asset creation and management goes, uh, you have the, at least the opportunity to create more efficiencies inside the organization.

Chris D’Agostino (33:47):

So let's shift on to some key challenges and yeah, Yao, I'd love to hear from you, you know, I just kind of teed up this idea that, you know, if you are working in a more federated way with data holdings in different systems, uh, you might have the, um, inefficiencies that I just spoke about versus with the idea of, you know, maybe a Lakehouse architecture or with the objective of just trying to get better collaboration between teams. Can you talk a little bit about, you know, what are the, what are the critical challenges that you're seeing to get different personas within the same team or department or business unit to, to work together more closely with data, and what steps are you taking to overcome that?

Yao Morin, JLL (34:31):

That's a great question. I think the answer to this question is quite related to the Lakehouse discussion. I think that number one thing that we did as we embark into this new data transformation technology transformation in JLL is to really have an architecture where, uh, in a federated model, everyone can contribute to the same on the same platform and, uh, contribute to that single source of truth. A lot of the difficulties and challenges in collaboration is the siloed of technology platform. And so the key, the key for us is to provide that platform and why now, where, you know, uh, data breaks, um, sequel analytics, and then the seamless integration with the lake, and then being able to assess, uh, different layers of data, including in, uh, you know, data warehouse and lake that really helps us to be unified on one platform.

Yao Morin, JLL (35:37):

And also on the, on the, on the other hand, you know, we in the data team, there are actually many different profiles and people with different skillsets and having that one platform that can accommodate different skillsets, uh, for example, analysts having more of a sequel skillset and then data scientists would more of like a Python or skillset. And then data engineering is really more about configuration of data pipeline or building data pipeline with Python that the single platform is really to help accommodate all these different persona and then be able to work together in a more collaborative way. And of course, we also want, uh, we also, uh, establish a really, uh, open contribution model where, um, the owners of different datas that is more about is more of a custodian of the datasets instead of like, that is my data is that you can't touch it, but really everyone can contribute with, uh, with the white process.

Chris D’Agostino (36:46):

Sounds great. Sherman, you know, part of your job title is data engagement. So we'd love to hear your, your thoughts on collaboration. And, you know, if you could also layer in a little bit about training and upskilling the talent within the organization, as some of these platforms come online and people are working with data in ways that perhaps they didn't in the past, uh, what steps has an organization your size taken to, to upskill the talent there?

Sherman Cooper, MasterCard, (37:14):

Yeah, so that's, that's a fantastic, fantastic question. And now had a great, uh, set of responses there, how to build on that. I think I would say that when it comes to collaboration, I think it's really important not just to collaborate, uh, as a data team and with our other data, uh, data teams across the company. So we're the data office, but obviously right, the business has its own data analysts, et cetera, but I think it's really having the collaboration and the, and the design of a data team that's integrated into the business. And so I sort of mentioned this earlier. So our data strategy team, which is a component of our data office really has what we call data strategy leaders that are actually embedded into each business unit, uh, across the company. And that is super important to create that relationship right, where we're understanding the needs of the business.

Sherman Cooper, MasterCard, (38:03):

And we're doing this process akin to privacy design, right? So we call it data design, right, where we are, uh, sort of piggybacking on the privacy by design notion, but, but really from the very inception of a new idea or a new efficiency in the business, the data team is at the table. One building that trust, but also understanding the implications of the, the, the needs of the business, right on, on our data strategy and how we have to be nimble in response to that. So I think it's really important to have a data team that's integrated into the business. So you have that relationship. Uh, but also I think it's important, you know, not to toot my own job here, but it's important. And one of the things that Joanne did, our chief data officer Joanne Stonier a couple of years ago when she created this role, she recognized the need, right?

Sherman Cooper, MasterCard, (38:48):

To have a sort of a data engagement lead, right. Reporting directly into the top of the data house. And that's super, super important because engagement is key. You know, one can have all of the, uh, you know, advanced technology in the world, right. But if the business isn't using it and the business doesn't see the value, and there's no use case that the business can apply the advanced technology to then it's, it's pointless. So I tend to describe this at MasterCard is like a tandem bike, right? So you ever gone on vacation, you see, you know, this, this, this long table everyone's drinking around the table, you know, having a few beers or having a few cocktails, uh, and then there are pedals underneath this thing, right? And there's some poor guy who's got to steer this thing right down the street.

Sherman Cooper, MasterCard, (39:32):

And no one's peddling. Well, oftentimes when we're thinking about data enabled collaboration, right? And you're in situations that are sort of like this, right? If you think about this as a metaphor. And so my job at MasterCard is sort of data, relationship, builder and cheif is to work closely with my colleagues across the enterprise, not just on the data team, but across the broader business, right? Understanding their needs, making sure that I'm working with our data literacy teams to ensure that the appropriate assets are out there for people to represent, understand the value of the tools that we have to the work that they're trying to do and to serving customers and making lives better, better for consumers and individuals at the end of the day. But it's also to do it in a way, right. That that is right side. Because oftentimes I think as data people, we can want the best tool, right.

Sherman Cooper, MasterCard, (40:20):

The most, um, the most cutting edge technology. Right. So just imagine that crazy looking tandem bike with everyone drinking and someone's steering, right. And then imagine me off to the left on a hoverboard, right. I probably can zoom down to the end of the street, you know, sort of defying gravity. But when I get to the end of the street, I'll be the only one there. Right. So even though I may have the most advanced technology, the journey is just going to be with me. And the key here is to get everyone to come along on that journey. And so it's figuring out how we steer the tandem bikes, but that means having the right relationships, understanding the use case, that's applicable understanding the current challenges of the business, but then also understanding where the business is trying to go. So after you build that credibility, as Randy was talking about on a couple of use cases, right, then you can start to map out where you're going to go in the future. And I think that's key to being successful in overcoming these collaboration challenges is building the trust initially having the right relationships and being able to balance solving today's problems with working on the next generation of solutions for tomorrow.

Chris D’Agostino (41:24):

Great. Patrick, I'd be curious about your point of view on this because, you know, you mentioned earlier in the conversation that you're kind of beyond the transformation stage now at McDonald's and you're using data in, in, in a very mature way. Tell, tell us a little bit about the personnel within the organization, how they interact between one another, how they interact with the data, like the concept of a single source of truth.

Patrick Baginski, McDonald’s (41:50):

Yeah. Um, so again, I, I have to agree here with Sherman it's, it's absolutely critical. And I named it McDonald's we, we call it the enablement, uh, portion of it, but the way that we really create our data science or our teams is, uh, we actually have a lead from the data perspective. And we also have a lead from the, from the business line. Right? So, so we actually create, uh, uh, collaboration at the team at the execution level. And the teams are very focused on the strategic priorities. So for example, delivery or digital, or, you know, the drive-thru and then those teams are really, you know, mixed in a sense, you have a couple of data scientists that come from the data organization. You, you may have data engineers from it, and you have, uh, two or three of the senior leaders in the business line actually right in the digital organization or in the restaurant operations organization.

Patrick Baginski, McDonald’s (42:47):

And they really work together on priorities and define the roadmap going forward and the data that they use and the tools which they're working off, they're essentially essentially strategized and defined, right? So we don't maintain multiple days data lakes or anything. We truly have one, you know, uh, repository essentially for our data. And we drive data governance from the top down, uh, throughout all of the markets. And we're really, you know, trying with this strong effort for collaboration across every business line to, to essentially, you know, continue driving good business value. That is 100% in line with the strategic priorities of the business.

Chris D’Agostino (43:31):

Yeah, that's great. So I went up, we only have a minute or two left here, so I'm going to just short circuit this down to, uh, Don and Randy, and with Don going first, what advice would you give to the people in the audience today? You know, you've done this at major league baseball, you're now doing it Northwestern mutual. What advice do you give to people who are perhaps aspiring to a CDO role? And then, uh, and then once you've answered Randy, if you can just layer in maybe some additional color onto Don's comments based on the research that you've done in the companies you've worked with.

Don Vu, Northwestern Mutual (44:05):

So, sure. Yeah. I we'll start with, for, with people that aspire to, maybe this advice also goes to executives that are trying to kind of transform the way data is leveraged in their organization. I think really a successful data strategy, like many other strategies comes down to people, process and technology. Um, so from a people perspective, the right leaders matter, and we've talked a lot about how cultural is the number one challenge and how both good and bad bad culture can be contagious. So really focused on getting the right leaders and not just at the top tier, but also those tiers below from a process perspective being bi-modal. I mean, I think it's, we need to deliver business value sooner rather than later. Um, and so I think as we're skating to the puck to build new capabilities, that will be innovative and revolutionary, we also need to deliver incrementally along the way so we can bring people along and really just show the validity of the path that we're taking. And then third technology, I would say that platforms can be enablers. So, you know, really do your best to pick the right platform, lean into them, treat them as paved roads that really help you on your journey.

Chris D’Agostino (45:09):

That's great. Thanks, Don. Randy final thoughts.

Randy Bean, NewVantage Partners (45:13):

Yeah, I would just say, first of all, change is never easy. We're talking about legacy companies with legacy systems and complex environments. So be patient or the role of the CDO is really transformational, but it's nascent. So organizations need to set realistic expectations of volumes of data continues to proliferate. Data is a journey you're never done. You need to continually continuously improve. That's the title of my book, fail fast, learn fast. So this notion of continuous improvement, and lastly, I strongly believe that data is the profession of the next generation. So there's plenty of opportunity and it's an exciting place to be.

Chris D’Agostino (45:53):

That's great. Well, Patrick, Don, uh, Randy, Sherman, Yao, appreciate your time today. I'm sure everybody appreciates it as well. And so thanks everyone for being part of the panel discussion.

Speaker 1 (46:06):

Thank you for joining this episode of champions of data and AI brought to you by Databricks. Thousands of data leaders rely on Databricks to simplify data and AI. So data teams can innovate faster and solve the world's toughest problems. Visit databricks.com to learn how data leaders are unlocking the true potential of all of their data.