Next-Gen Customer 360 With Data + AI

Read Interview

Chris D’Agostino, Host:
Welcome to the champions of data and AI series. I’m Chris, D’Agostino your host for today’s episode. Don BU chief data officer at Northwestern mutual tool is back to discuss something near and dear to every organization, how to be customer obsessed. In other words, how creating a 360 view of your customer can help improve all things related to the customer experience and increase customer loyalty. In this episode, Don shares how his personal experiences with data and AI are helping him take customer centricity to the next level. Dawn, welcome back.

Don Vu, CDO, Northwestern Mutual:
Thanks for having me today, Chris. Good to see you again.

Chris D’Agostino, Host:
Yeah, it’s great to see you Don. Hey, so, you know, we we’ve talked several times now and I want to get started with your background and how you got involved in data and AI and programming. Cause it’s really sort of a fascinating story. So if you can take us back, you told me about when you were six and kind of your first exposure to programming, which, certainly given where we are in our careers starting a program at six years old was pretty unique experience and unique opportunity. So how did that shape you know, just your outlook on data, AI programming?

Don Vu, CDO, Northwestern Mutual:
Yeah, I guess I can consider myself pretty lucky or maybe a little bit spoiled. So when I was six for Christmas, my parents were amazing. I got my first robot. It was a, it was called the Omni bot 2000. I think if you can Google it, you’ll see a couple of cool images come back. I recall getting in and being like a kid in a candy store, I think that you could program it with cassette tapes, but that was one of my first exposures to kind of just technology and just the notion of actually communicating with machines via programming. And soon after that I actually got put into computer camp. So I think it was around sixth grade where I ended up going to a summer school where we did computer camp. Each of us had, we were all part of different teams in the class.

I was part of the microchips and that’s where we learned about, or that’s where I first learned about logo logos a programming language and early one where we would have a turtle like graphical turtle that we would actually control. We would have moved forward 50, right turn 90 and actually draw some shapes. We would put a pen down and it would actually draw shapes graphically on our screen. And we actually had a physical robot as well. So we had a physical robot that would draw on the floor of the pen would go down and withdraw all sorts of shapes. I’d have four loops things that nature and actually ended up drawing. And it’s not until later on that I found out that logo, one of the co-inventors is a gentleman named see more pet pear. He’s also a pioneer of artificial intelligence. You worked on both while he was at MIT. So now like working on AI initiatives and having been in data for well over 20 years for my career, it’s just kind of funny and almost full circle, how I started with logo working in AI now and almost full circle with the Seymour pet pair there along the way.

Chris D’Agostino, Host:
Yeah. Awesome. So, you know, you’ve had a great career, you’ve been at WEWORK, Major League Baseball now at Northwestern Mutual. What are some of the consistent things that you’ve seen in terms of your passion for programming and for data that have kind of helped you shape this career?

Don Vu, CDO, Northwestern Mutual:
It’s so interesting. I think everything kind of makes sense when you look back at it. But maybe along the way and the way it organically manifested it wasn’t exactly how it was planned. When I first came out of school, it was really a data engineer to start. Soon after that started getting to work like a CRM sort of perspective. So understanding how to leverage that data when it’s aggregated and with a customer centric lens later on got very specific about leveraging customer data for analytics specifically for baseball for 13 years. And it was just a, it’s just a natural progression.

Chris D’Agostino, Host:
When you think about customer 360 has different meanings and, and certainly in, in a particular vertical, it has a different meaning. We’d love to hear what your general thoughts are on customer 360. What comes to mind when, when you hear that phrase and how do you apply the lens of, you know, insurance versus major league baseball and the different verticals you’ve been in?

Don Vu, CDO, Northwestern Mutual:
Yeah, no. When I hear clients 360, I basically think of the phrase, whatever your company wants. I mean pretty much every consumer business aspires to have this 360 degree view of their customers, every touch point, digital and offline aggregated in one holistic view. And they all want it for the same general reasons. They want to understand their customers. They want to understand by the very segments that matter for them and ultimately the segment of one each individual. And so when we were at baseball, you know, certainly leaders at baseball won’t understand how fans are engaging with their tickets, engaging with the stadiums and then how they’re engaging digitally and online at a place like a, we work folks wanna understand how customers are engaging with the physical space that they’re actually occupying and at a financial services company like NM that has a unique value proposition that combines risk and insurance products alongside investments.

That holistic view is just as important. Every company wants to understand these, all of these touch points and really what they mean. And the reason for that is just to have a deeper relationship to, to serve folks better to have a long lasting relationship and have these serendipitous interactions, amazing things about the challenges of data and all their solutions. You have these incredibly different industries, you’ve got baseball, financial services, real estate, but they all have similar problems. And that’s why I said a couple of times, but the problems and their solutions and data aren’t the same, but they tend to rhyme. You know, you shared one example and I would call it just customer centricity most simply. But really it’s, there’s, there’s a bunch of other ones. There’s just the power of platforms and unified data to address silos of data. There’s the tight synergy that’s required between business strategy and data strategy in order for there to be mutual success. So it’s been neat to see over my career how, again, the similar themes all along the way.

Chris D’Agostino, Host:
Yeah. There’s a bunch of different touch points that customers have when interacting with an organization. So if you take major league baseball and you talk about tickets and then a stadium interactions and then the mobile app or the website and then when you’re looking at Northwestern mutual policies and you know, probably incidences that get reported and things like that help me understand. So it’s kind of an omni-channel environment, which means you’re pulling data in from a lot of different sources. How does, how does the architecture that you talked about for that unified platform? How does that, do you see that being very similar between these different verticals in your experience and how do you pull all that data together and create that view?

Don Vu, CDO, Northwestern Mutual:
Yeah, and I think there’s like a lot of similarities and there are certainly some differences at the end of the day, the data architecture and the technology that powers, it is really so critical because aggregating this customer data from like a wide variety of sources with different levels of velocity, different levels of crane is challenging. And so technology, I feel that we’re very much in kind of a golden age of technology with respect to data. So I think we’re pretty fortunate to have a lot of the tools at our disposal. Data architecture is critically important and making sure that you have the right tools to get all that data in the right place. I think some of the ways that things may be different than a things like data latency and the velocity of data in each of these various contexts, certainly dealing with videos, streaming, telemetry, like we had to deal with it MLB with things like mlb.tv or video streaming platform, that’s going to, you know, have a different challenge than some other different types of data. So at Northwestern mutual, we have touch points with our financial representatives and our various clients. We have our digital apps for which we have tons of telemetry as well. And then we have them calling like service that our home office service folks. And so there’s that whole call center sort of interaction as well. So I would say there’s a lot of complexity and a wide variety of types of data. So that requires an architecture that can accommodate all of that and just wrangle it all together.

Chris D’Agostino, Host:
Yeah. So the speed at which the data arrives the volume and then the time that it takes to actually execute the use case that that you’re trying to build against that data. We did, you know, certainly in my previous job, we had a huge infrastructure around call center support for the bank, and we had telemetry on the app and the mobile app and seeing, you know, customer interactions. And, and of course we created an AI based, a digital assistant to try and one improve the customer experience, but then also drive down some costs inside the infrastructure of the company, because we had a lot of call center agents and we wanted to see what we could do to reduce that. I know from, we talked to, you know, you talked about sort of a personal experience of you know, having to go through or looking at that customer 360 viewpoint and you know, the, the pin cushion comment about blood work, getting done, wanted to see if you’d share that with the audience and talk about how that’s shaped your thinking a bit.

Don Vu, CDO, Northwestern Mutual:
Yeah, no, I think it’s, it’s interesting. I think whenever you’re taking on a role such as chief data officer or a similar data leader or any leader, really in a company, I think you certainly have a certain inside out perspective, but then you also try to put yourself in, see to the customer. And then for me having come to Northwestern mutual and actually being a client before I arrived, I think like many new fathers, when my daughter was born, she’s, she’s going to be nine this year when she was first born, I felt certainly a responsibility to make sure she was taken care of no matter what. I ended up getting life insurance and lo and behold, coincidentally, I ended up going to Northwestern mutual for it. And again, this is long before my employment here. And it was interesting going through the underwriting process.

Don Vu, CDO, Northwestern Mutual:
I think like many life insurance processes, it’s really relatively common. You have to fill out a bunch of information. You go through the process of getting your fluids drawn and there’s a risk assessment process that goes on. I had a little bit of a challenge when the, the gentleman was trying to find my vein and that you know, we had to try both arms and it took a little bit of time and hence the pin, the pin cushion analogy when we got there. And it it’s it’s for good reason. I understand why obviously certainly being on this side of the world and it’s, it’s all of that information. And ultimately data that’s acquired during that process that allows companies like Northwestern mutual to make a very robust risk assessment. And so it’s been interesting to see now in the seat that I’m in, as we try to innovate on that process with the customer experience in mind, how do we make the process smoother?

Don Vu, CDO, Northwestern Mutual:
How do we make it less invasive? How do we streamline it? And how do we shorten the time from the beginning of the application to actually having a policy issued has been really neat to try to tackle that problem using AI and machine learning, whether it’s AI and computer vision, to understand things like attending physicians notes, whether it’s actual machine learning for the risk assessment piece. It’s been really, really neat to see that all evolve. And again, having been on the other side of the, on the other side of the wall for it

Chris D’Agostino, Host:
Well, and so you’ve got a better appreciation for having been through the experience and now you can help lead some of the changes around it. And then obviously the last 12 to 18 months have been an interesting period what to find out like, what are you doing to react to maybe fewer physical touch points with customers at this stage of the game?

Don Vu, CDO, Northwestern Mutual:
Yeah, I think like a lot of folks, it’s funny, as soon as that meme that was going around for quite a while, like what has been driving transformation and, and there’s, there’s been a joke. I think that the last year there’s been incredible transformation. That’s been driven just by circumstance and not necessarily any one individual. And I would say for us as well, like there was a tremendous acceleration of a lot of like the digital niches that we’ve had before and really an embrace of those initiatives. We feel very fortunate to have had a lot of tailwinds because there were, we had a lot of momentum already in place. It just accelerated it. So really an embrace and a ramp up of those digital touchpoints those methods by which we actually collect information such as our online medic medical health questionnaire. And again, the risk assessment, leveraging machine learning and other accelerated underwriting processes that leverage all that data. We’ve really seen a dramatic increase in that. And we’ve actually doubled down. We’ve really responded to the incredible productivity of our field representatives and the high demand for our product, and really to ensure that we continue to serve our clients as they’re needed. And we’ve really doubled down to ramp up on that.

Chris D’Agostino, Host:
What have been some of the key challenges or pitfalls that you’ve seen either just sort of in this experience of getting that 360 view of the customer in light of everything kind of going on in the world or, or in some of your previous roles? I was with a customer yesterday and spent a full full day with them and was actually nice because it was in person and we had, you know, breakout sessions and brainstorming. And we were talking about just the challenges around getting the different data assets that they need to create the intelligence that they want associated with the use cases wanted to see, you know, what things you’ve seen as challenges in the last 12 months or so with trying to, trying to get this data together and, and get that insight.

Don Vu, CDO, Northwestern Mutual:
I think it’s fair to call out the technical challenges. I think like many folks in it’s, it feels cliche that this is hard. But having been at a bunch of places, it can’t be understated is that it is a hard technical challenge to get data from all these disparate sources to merge them all together, to be able to join all these silos. I actually feel like there’s a ton of other like cultural and business challenges. And when you think about even just like, what is, who owns the customer in an organization, right. I think many organizations have lots of verticals. Many of those verticals are, most of them ended up touching a customer. So even the notion of who owns the customer, it can oftentimes be something that’s up for debate. So what that means is that you’re aggregating this 360 degree customer view, and you’re trying to gain enterprise alignment.

There’s a lot of, you know, relationships that need to be maintained. There’s a lot of alignment and share that needs to be bought into. So we really need buy-in from the top in order to make sure that folks know that this is an enterprise wide initiative, this is critical for the company. And once you have that, I think you can then start to get people to coalesce, to really get and rally around this notion, because everyone’s going to want to use it as soon as you have a client 360 or customer 360 view, it’s something that’s gonna be relevant to almost every part of the business. So for me, I would say that there’s just as many business challenges as there are technical.

Chris D’Agostino, Host:
Yeah, that’s it that’s really consistent with, with what I’m hearing, talking to other leaders, especially in a multi business unit organization where the definition of a customer might vary slightly, and then the definition of a product that you’re offering to a customer creating that common sort of language or ontology. And, you know, we’ve, I’ve never been a fan of, you know, big, you know, enterprise data definitions. And, you know, the schema is to sort of schema to rule all the data. I just think those take forever those exercises to try and create that single view. However, creating kind of a very lightweight ontology of some of the key entities that you’re trying to track within an organization is something that we’re seeing take shape quite a bit. So just wondering, like, you know, have you looked across you know, the different business units and looked at like, you know, what are the key elements of a customer definition and then how do those even relate to maybe CCPA or GDPR type regulation?

Don Vu, CDO, Northwestern Mutual:
Yeah, no, I think you touched and an interesting theme and it’s almost like the notion of centralization versus decentralization, and there’s a lot to wrangle. If you want to have everything funnel through one explicit canonical data model. I do think there’s a benefit for, as you mentioned, like certain key entities to have no real dissension on what that is and fragmentation. So I think the customer is one of them. So like having a canonical customer entity makes a ton of sense leveraging perhaps like mastering and the old school MBM has one method by which you break down silos. I think every organization’s a little bit different, but in most, in many cases you have multiple systems with a different view of, let’s say Don view, who is a customer, how do you actually reconcile that? I think that activity when done for the good of the entire enterprise and having one consistent view such that Don VU at when he comes into these various touch points has a consistent experience. I think that can be beneficial, but I think it’s a fair point. I think that I don’t know if that scales across every single entity and entire enterprise. So I think there’s a nice balance that needs to be played between the two. Yeah, I think it’s, it’s it’s the challenge that I’ve seen it pretty much at every single organization I’ve been in. It’s a, it’s a tough nut to crack,

Chris D’Agostino, Host:
But we’re doing yeah, well, this and this sort of ties into the thing that we talked about at the panel discussion, which is sort of the cultural elements of the business processes that the adoption and buy-in, and just sort of the philosophical buy into the efforts and, you know, the enterprise approach to, to try and to do this customer 360. I know you created a data strategy steering committee. Can you tell us a little bit about that? What’s the composition of it? What was the motivation for it and what do you leverage that steering committee for?

Don Vu, CDO, Northwestern Mutual:
Yeah, I think it’s funny. Like we stood it up for a bunch of different reasons and it actually touches on what I shared before, what I think some of the biggest challenges aren’t we talked about the technical ones, but really like the business challenges and getting that mind share and that alignment it’s funny, I think particularly given the circumstances of the last year and most people being very remote and not be able to get into a room, having these forums that were intentionally created to try to create alignment, to cry, to try to create mind, share, have been really, really instrumental in our ongoing success. So when we first formed our data strategy, executive steering committee with representatives from all across the business and then privacy and enterprise architecture all across the org we really stood it up such that we could get feedback and as quickly as possible as we actually had a roadmap and spun it up as we started to deliver on it, as we started to deliver business use cases, get our partners to pressure test what was being delivered.

Keep us honest. It’s been really critical for us to make sure that we’re not deviating too far. It’s not always a straight line from a to B, but it’s not nice not to get too off course. So that’s been been great. And then from a almost cultural perspective, it’s less, let’s let us collectively have a shared idea of what the art of the possible might be and then how it actually manifests in reality. So we might talk about things in concept, but then as we highlight our success stories, as we present with our partners in the business, things that we’ve achieved over the last period of time, it’s actually cool for their peers, business, executive peers, to see what other departments are doing with respect to the AI and machine learning. So it’s been a really cool collective journey and essentially it’s almost a flywheel. Once you start letting people understand how you can unlock the power of data in certain ways that generates more ideas, and then it just kind of like feeds into itself. So it’s been a really beneficial for the organization.

Chris D’Agostino, Host:
And how, tell us about your role relative to the steering committee. Are you a member of the committee or is the committee someone you present to, for example, we had a data steering committee at capital one when I was there and myself, the lead for definition of product and the lead for the user experience sort of all co presented to the steering committee and the steering committee was made up of members from across the lines of business, as well as the, in tech and the objective there for us, the steering committee was really someone, a group that was going to help us ensure that we were focused on the right priorities as the priorities needed to shift based on business circumstances or technology issues that we would run into. We wanted to always have this sort of check back with the steering committee to say, okay, here’s where we’re at with the transformation. Here’s where we’re running into impediments, here’s where we’ve got budget risk. So they actually, you know, we’re sort of taking down obstacles for us and making sure that we had, frankly, some top cover around some of the decisions we made as we did the transformation. Is it similar with, with how you’ve set this up or have you, you know, we’re using a different model?

Don Vu, CDO, Northwestern Mutual:
Yeah, a little bit different. I would say there, these folks are closer probably to the delivery and the implementation of it. And the business use cases. We have another forum through which we actually unblocked some of like the other budget issues or something, things that need to get elevated. This one is the folks who are a little closer to the eye. And so it’s almost like a co-creation and we’re Mo we’re going on the journey together with these folks. But I think it’s a really good call out that you have to make sure that you have those avenues by which you can kind of unblock some of these bigger issues. And we’ve been really fortunate to have like a ton of great support at the top executive levels to do that. But on the whole, yeah, I would say that like, we’re, I actually really appreciate the partnership that we had in the collaboration how we go hand in hand between our data teams and our business partners.

Chris D’Agostino, Host:
And can you talk a little bit about your personal role in terms of articulating the strategy and educating, you know, the senior stakeholders, as well as the people that actually have to adopt and use the system and platform on a daily basis? Like tell us how much of a, you know, sort of PR champion you ha you’ve been within the organization to, to help get this thing going,

Don Vu, CDO, Northwestern Mutual:
You know, quite a bit a bit. I mean, I shoot, I remember when I first got there, one of the first things I was tasked with was presenting to our board just to like early impressions and what’s our data strategy. And I do feel like in forming that and presenting to them, and then even presenting to the folks throughout an M at the leadership level and on the ground, as well as quite very much a road show. And it’s something that I’ve really taken the responsibility to be an evangelist, right. At the end of the day, you would certainly want to have folks understand and have buy-in for your initiatives. I think that’s always been kind of part of the gig wherever I’ve ever been. So I’ve been getting involved at the most granular levels and actually like solutioning and getting our hands dirty with how we’re actually going to get stuff done. Again, I’ve been really, really fortunate to have seen a lot of stuff over the years in my various stops. But also trying to like keep it strategic too. I think oftentimes like leaders, we have to like operate at a bunch of different altitudes. You’ve got to get all the way into the weeds and then you got to level it up quite a bit as well. So that’s part of the fun of it.

Chris D’Agostino, Host:
Okay, Don. So you’ve said like, you know, you’ve established this steering committee you’ve got engagement from senior leaders in the org organization. You’re serving as an evangelist to, to make sure that adoption and buy-in is, is widespread. Let’s talk a little bit about the actual implementation for a customer 360 and how that has evolved over, you know, from an architecture standpoint, from an implementation standpoint, how that’s evolved over the different stops that you’ve had along the way with your career.

Don Vu, CDO, Northwestern Mutual:
Yeah. And funny, and in two decades, plus to see how things have really evolved in the data space. And like I said before, I feel like we’re in a bit, a golden age of a lot of that technology. I recall like earlier in my career, when we were doing analytics kind of data warehousing sort of efforts, creating star schemas and having almost data warehouse architecture within an oil LTP system, like an Oracle sort of system, and certainly saw the strains of that and then went to a more, more purely analytical system. It was an MPP sort of situation. I think there are folks that have been in the business for a while, probably recall like the teases and the Terra data’s of the world. And then soon after that and moving to things like Redshift and Presto, and then even snowflake and where we are now and what we’ve been really embracing at Northwestern mutual.

Don Vu, CDO, Northwestern Mutual:
And as we create what we call our unified data platform, one place within which we have all of our data for both descriptive analytics and also predictive analytics, we really wanted to have an ecosystem that was a governed self-service system. We have all sorts of data ingest tools that were spinning up such that application and technology users can be, can take ownership of getting their data into the platform. The platform itself is built on Delta lakes. So it’s an open source product that again, allows us to have the sensibility and all the functionality that we need there, and then leveraging other tools like Presto and the consumption layer, but really like focusing on this notion of like a, a lake house architecture. What I really like about that, having seen some of the constraints of the traditional EDW, even with some of the more modern implementations, is that we can run, you know, machine learning workloads directly on top of this leg. You could spin up a spark cluster and it could sit right on top of that same Delta lake ecosystem, the same one that you might use for acquiring using a BI tool. So that’s really, really important to me. And I think that’s a nice evolution. You don’t have to worry about synchronization between different environments for these various use cases. You have a lot of elastic compute. So it’s super exciting to see the way the whole field has evolved in a really embracing lake house.

Chris D’Agostino, Host:
So Don, we always wrap each episode asking leaders like yourself, what advice you would give to other aspiring leaders and in people that really are pursuing a career in data and AI and in a, in a more senior role. So what’s top of mind for you.

Don Vu, CDO, Northwestern Mutual:
Yeah. And I guess a couple of things, the first is embrace being a lifelong learner. Obviously the field changes quite a bit. I didn’t study machine learning when I was in college. Certainly had to pick it up up along the way, and there’s going to be more things that I need to pick up in the years to come. So just embrace that. And it’s an, it’s a great opportunity to continue to just use different muscles and just, just accept it. And as part of the journey the second thing I would say is that relationships matter at the end of the day. So much of what we do with data, given the fact that it’s so horizontal in nature, given the fact that the tentacles, you know, get into all the nooks and crannies of the organization, you’re going to have to deal with folks from all the all stripes from all over the organization. So embrace that, embrace that the fact that EKU relationships matters, building consensus matters. And yeah, it’s all just part of the journey. And I think if those two things are, are you just take those into account and enjoy the ride with those, then there’ll be better off.