The CDO to CDO podcast is hosted by Chris Knerr, Chief Digital Officer of Syniti.
Subscribe Here: Listen to all the episodes in the Syniti Podcast Series
Chris Knerr (00:08):
Hi. Welcome to the CDO Magazine Interview series. I'm Chris Knerr, Chief Digital Officer of Syniti, a world leader in enterprise data software. We're partnering with CDO Magazine, MIT, CDOIQ, and the International Society of Chief Data Officers to bring you the series of interviews with thought leaders in data and analytics. I have the great pleasure today of talking with Danielle Crop, Chief Data Officer at American Express and a CDO Magazine 2020 Global Data Power Woman, selected by CDO Magazine for being a pioneer and a key influencer in shaping the data and analytics industry. Welcome Danielle, great to have you on the podcast and terrific to meet you.
Danielle Crop (00:47):
Thanks for having me, Chris.
Chris Knerr (00:50):
Absolutely. I thought we could start out if you would, by just giving us kind of a brief career sketch, how you came to be a CDO, which I understand is an enterprise role within American Express, were there are a couple of pivotal roles or experiences and then sort of a brief outline of the scope and accountabilities of your group?
Danielle Crop (01:13):
Yeah. Sure. Of course. I have a bit of a unique pathway, I think to the chief data officer role. I actually have an undergraduate degree in, of all things, forestry, quantitative ecology actually. I wanted to make better decisions about the environment, so I went on and got a master's degree in environmental statistics. I thought my role in the world would be to count birds in the forest. That did not happen. I had to stay put after graduating, my husband was finishing his undergrad program in computer science, and got a job at a small Subprime credit card company, and really got very interested in all of the data and all of the challenges, and all of the problems that could be solved with data in the financial services' industry. Then I actually went and deliberately looked for a job at American Express because if I was going to be in financial services, I wanted to work for the best.
As a customer, I had had my best experiences at American Express, so I chose to find a job at American Express. I've been here for quite some time now, but a couple of the pivotal roles at American Express that I've had that really I think prepared me for the CDO role was one, a role that I had in fraud risk management. That was a role that included all of the data and analytics and modeling skills that I learned as a statistician. But also helped me to learn about the technology side and the data aspects of solving, how those two things come together to solve problems because in fraud, if you don't catch it quickly, you don't catch it at all. It's not just about the data, it's about the systems and the delivery of those decisions etc, in a very real time environment.
Working in that space really got me, my first real job in what turned out later to be software product management, right? But at the time was just, we were thinking more of as project management in the early odds when I did that role. I learned a ton there. The second role was when I started leading the digital acquisition space at American Express, and we went through agile delivery transformation. In that space, we really needed to come together and cut across silos, bring teams together in order to deliver end to end in a scaled agile universe. Through that experience, I learned a lot about change management, a lot about people as systems and how to really bring those things together to affect change across an organization and across organizational silos, which I think is really what the Chief Data Officer role is a lot about, is cutting across those silos and making a difference and making the changes.
As far as my role as CDO and what it entails at American Express, the vision that I have is to realize the potential of our unique data assets, to power the world's best customer experience, right? Which goes right along with our vision of customer experience as a company. In that role, I'm accountable for consistent, usable and trusted data across the enterprise, which includes the data management aspect of that role and doubling down on our successful use of data in the company over many, many years now in the areas of marketing, risk and servicing and ensuring that the company uses data in a way that it really brings the trust of our customers as a key aspect and ensures that we are complying with all the changing regulatory landscape around the world these days. That's really the function of the role.
Chris Knerr (05:22):
That's just really interesting about your background too. I have an undergraduate degree in philosophy and ended up doing all this digital and technology stuff. It just goes to show, wherever you start there's a lot of pull towards all this interesting work. I really appreciate your observation too. How much of digital data transformation roles are about change management within an organization and then, very interesting what you're doing with data, obviously, kind of a tremendous heritage and being kind of an early adopter from an American Express standpoint of managing all that customer data, which I think will be a great opportunity to explore in our conversation. Maybe we can kind of double click on the customer piece of this for a few minutes.
When I was thinking about American Express, there's an argument to me to be made that American Express maybe was the original platform company before there was such a thing as a technology platform company. I see that one of your key strategic pillars is quote, making American Express an essential part of our customers digital lives. Just kind of tying back to world-class customer experience, who do you think about as being your customers and kind of how does that inform the strategy?
Danielle Crop (06:54):
Yeah. We have a very broad base of customers, right? Our customers go from consumers to merchants, to small businesses, to very large multinational corporations. Across that different customer universe, it's really important for us to bring all that data together in a way that helps solve customer problems, right? One of the ways that we do that is through something we call Customer 360, that capability. Yeah. Go ahead, Chris.
Chris Knerr (07:27):
No, I mean, maybe I can just drill under that. I've often felt like in thinking about the digital space and the data space that business to consumer is kind of the easiest model for people to understand and sometimes companies trying to do data work and digital transformation, try to paint the whole world with that brush of B2C. In fact, you're in a unique position as you articulate that you've got kind of three very different channels or major segments. Maybe amplify a bit some of the differences in the strategy from a customer experience standpoint, and then how those are brought to life with your data assets.
Danielle Crop (08:08):
Yeah. That's a great question. I think that going back to the Customer 360 capability, I think it highlights what you're asking very well, which is that we have one single capability that brings all of our data together across all of our different lines of business in order to create a complete view of the customer relationships that we have. By doing that, we're actually then able to reuse that capability across all sorts of different experiences. Let me give you a few examples.
One is that if we have a sales experience in our corporate or small business environment, in which we know we have a consumer relationship with that person, we're able to bring all that data to bear in that sales experience and customize that experience for that relationship we have with them. We also have the other aspect which is, we understand whether or not a merchant has a corporate card relationship with us or they do not, and we can then customize that experience based on that knowledge and that understanding of, how much business are we doing with this customer and how deep is our relationship with them. I think that that can make a huge difference in the way that we interact with each of them.
Chris Knerr (09:35):
If I understand your point correctly, and this is very interesting. In a sense, you could think about, there's one sort of view that the channels and the segments are rather different but part of the magic is that because you're actually interacting, if you look at it from kind of the customer experience out rather than the segment in, you can bring to bear different facets of the data from an end customer standpoint and merchant standpoint in order to provide the best possible experience. That's interesting and I think it speaks to both the unique characteristics and unique assets that you have as a company, but if I'm following correctly.
Danielle Crop (10:17):
Chris Knerr (10:20):
In that vein, so the very interesting like chief data officer, right? I actually had all three of those experiences with Amex as a large corporate customer, as an SMB and then as an individual consumer. One of the things, they're three companies in the world, I'm exaggerating, that still have real customer service. Old-fashioned traditional customer service, if I have a problem, I can call up and speak to a human for which I would say, "Thank you." I love that about American Express. But in a sense, it's almost like a non-traditional approach now because the trend has been to just outsource and outsource and kind of push customer service back onto the customer rather than viewing it as kind of an asset or a capability that's strategically valuable. In that sort of data realm, intersecting segments that we talked about, and then the idea of traditional customer service, what's the interlock? What's the benefit of maintaining both of those and that kind of idea of your customer's digital lives?
Danielle Crop (11:33):
Yeah. We see them as complimentary to each other, right? Versus being a trade off. What we try to do is we build digital experiences that are based on data like Customer 360, that help give the right experience to the right person at the right point in time, based on the data, the machine learning and the design of the experience, all working together in concert. Then when necessary, we can direct people when touch points are challenging, to reach out to a customer service representative to get the type of experience they're expecting in that particular interaction. We see these as all very complimentary channels to each other in delivering the best customer experience.
Chris Knerr (12:20):
I agree conceptually but in my work, I deal with a lot of customers that are kind of chasing cost optimization and I think that there's a view that it's sort of, you're saying it's a positive-sum game. I think most companies see it at best a zero-sum game and possibly a negative sum game. Has there ever been any pressure to sort of like, "Well, we have this, it's expensive. It's valuable but maybe it's too expensive. Let's push more and more to automation." Or is this something that you would sort of view as in a way like a crown Juul of your market position and differentiation? What do you know that your competitors don't know? Who are all doing this and to me, what's kind of it's a bad way from a customer standpoint?
Danielle Crop (13:16):
I think that we have always put the customer experience first, right? That's something that is inherent in our brand. It's inherent in who we are as a company. I think that by taking that customer first perspective like you saw with Customer 360 and seeing relationships across, and by seeing interactions across all channels and understanding what is the best channel to do what interaction in, for the customer experience. Of course, we have automation goals like any other company.
Chris Knerr (13:16):
Danielle Crop (13:47):
But we want to use those in the service of balanced outcomes, that we want great outcomes for our customers. We want to retain JD Power. We want to do those great things, but we also want to be great for our shareholders as well. We have those balancing aspects of how we make decisions in American Express.
Chris Knerr (14:10):
Yeah. No, that makes sense. One is that, and just on a personal level, I agree there's part of the human touch that just can't be replaced when you really need it. But then there's also, I think from a data and data science standpoint, if I follow your line of thinking, in the same way that the best chess player in the world isn't a human and isn't a machine. It's a human guided by a machine. That's maybe a good analogy for your philosophy of customer service that all that automation and the 360 degree data and the machine learning, fraud prediction. But at the end of the day, you want to actually have the human touch and human intention kind of guiding that customer experience so that it really isn't just purely automated.
Danielle Crop (15:00):
Yeah. I think a great example of that is our messaging within our American Express app. There is, when you go in there and ask us a question, of course, there's automation and there's AI in there, but there's also very good intelligence around. Once you get to a point where the AI can no longer serve you, it goes seamlessly to a customer service professional.
Chris Knerr (15:26):
Right. If you're able to share without getting too far into proprietary stuff, maybe I'll ask the question this way, how have you designed the system to sort of know what the tipping point is, like when automation's not working, because on the customer experience side, if you're the customer of that, you can tell instantly and it becomes exasperating very quickly, right? You must have done something smart to kind of design that customer experience system to try to identify sort of trigger points or tipping points which is an interesting balance, I think of data work, AI technology work, and then kind of change management and business philosophy.
Danielle Crop (16:14):
Yeah. It is bringing all those disciplines together, right? Is the key, right? Understanding when the decision science, like where is that going to work and doing the analytics around where that's going to work, and then where it's not going to work, and then designing the systems around the data-driven analytics, I think is a key to the success of the things that we've done.
Chris Knerr (16:45):
Yeah. Makes sense. You kind of answered this already, but maybe just to sort of put a punctuation mark, if you were to characterize for the audience, is your role kind of more about growth and differentiation or more about efficiency and optimization and cost savings, or what's the balance that you strike in terms of managing those data assets and kind of figuring out how you contribute to driving the strategy?
Danielle Crop (17:18):
I don't think they're mutually exclusive, right? I think they're all of the above. But I think that there're ways in which you can drive efficiency through innovation and vice versa. My role in all of these different things is realizing the potential of the data assets, whether that be for efficiency or whether that be for innovation, very much a matter of the use case, right? In this scenario. I think that we're attempting to drive both outcomes.
Chris Knerr (17:54):
Yeah. Well, no, of course, and I didn't mean to suggest that they were sort of binary opposites but maybe just to drill into that a little bit. I mean, when you were speaking about your background, I have the impression that your role is kind of a combination of senior level product management, right? But then also business strategy. Are there a couple of areas in which you've used data assets to drive growth or to drive differentiation, maybe outside the immediate customer experience area that you could share? The reason I'm asking this is because I sometimes feel like the whole AI data asset world is overly focused on cost optimization at the expense of growth. I agree with you philosophically, I don't think there's necessarily a trade off, but I'm always very interested in where companies other than pure play data companies are using their data assets to drive growth or to drive differentiation.
Danielle Crop (19:01):
Okay. I think that it's hard to pinpoint any specific because it's actually part of the DNA, right? The data science underpins so much of our products in so many different ways. Let me give you an example of one of the aspects that we do is experimental design, right? Actually within our digital products, so we will consistently be testing and learning on the digital products to understand what created the most seamless and easy experience for a particular consumer and by doing that work and it's literally in the day-to-day work of the digital products that we have. It gives you a sense of how, I think the combination power of what we put together as a combination of the digital assets and the data and using them together to power the experience.
Chris Knerr (19:55):
Yeah. No, that makes sense. Thank you. Maybe shifting gears a little bit. Let's go into sort of data ethics, AI ethics, and regulatory for a bit. Can you share at a macro level, how important is the regulatory landscape either kind of backward looking or forward looking in shaping and driving the direction of the data market?
Danielle Crop (20:24):
Yeah. I mean, as a financial services institutions, it's highly regulated. The regulators are a key constituent of mine as the CDO, as well as the consumers, right? All of our customers. I see the regulators as a key constituent and we are consistently looking at all the regulation around the world and making sure that our products and our experiences and our data assets are complying with those regulations around the world. It's key to us and it does drive a lot of the aspects of how we go about managing our data assets. But I think that that is clearly in many cases, inline with what our customers expect of us, right? Which gets to your point around ethics, right? I think that we are working right now on really putting together internally, a set of data ethics that will cascade across the enterprise, so that we're all very clear as an organization around the ethical principles and having our customers' backs, as well as meeting the regulatory expectations.
Chris Knerr (21:42):
Yeah. No, that's very interesting. I guess to just drill into that a little bit, I'm wondering, my view is that to some extent, the pace of technology change has been so rapid over the past 10 years in particular, that it's extremely difficult for the regulators to keep up. If you look at an issue, which I'm sure is top of mind and I'm sure you have some good thinking about on for example, algorithmic bias, right?
Algorithmic bias for those in the audience who aren't familiar with this and particular in machine learning is, it can be sensitive to what data the models are trained on. When you're dealing with individuals data, you have to be very careful that economic bias or demographic bias doesn't creep into the predictive models. My sense is that the regulatory community hasn't caught up with that yet and so, I'm wondering if you think that market-leading companies like American Express have kind of a role to play, not just obviously in following the regulations but in providing thought leadership to regulators to kind of help make sure that they don't fall too far behind the fantastic exponential pace of technology change that we have been experiencing and will continue to experience?
Danielle Crop (23:09):
Yeah. I mean, I do think we have a role to play, right? In educating and in influencing policy makers etc, on these particular areas. I do think that we are current as far as algorithmic bias like the company, we are definitely working on that at the moment and we will continue to do so. It's part of our values, right? As a company, to have our customers' back and to do the right thing. These are just part of Amex. We want to make sure that we are making decisions that are as unbiased as they can possibly be. That is what we're going to be working on as part of our data ethics program.
Chris Knerr (23:53):
Yeah. No, that's great to hear. I wonder if you, again, this is just my point of view. I have sort of an emerging view that there are a couple of industries that have kind of a special role to play in helping to advance that regulatory agenda in a way that is thoughtful. Not crippling growth but not sort of letting, making sure that consumers and businesses are protected. One is, I have a life sciences background, so I tend to think of medical data. PII medical data is playing a special role but I kind of see the same thing in financial services, that because of the privacy and how critical that data is to the lives of humans, that your industry has kind of a more key role to play and a leadership seat at the table in terms of kind of advancing that regulatory agenda in a balanced way.
Danielle Crop (24:53):
Yeah. Absolutely. I mean, we feel that our customers have chosen to trust us with their data and to protect it and it's an important aspect of how we think about our data.
Chris Knerr (25:06):
Yeah. Super. Shifting gears again, another area that I'm very interested in and I value your perspective on is, one of the top issues that I hear about in the industry in particular from Chief Data Officers, Chief Digital Officers, and CIO is around talent. What's your sense of kind of the talent pipeline, are we on track? You can sort of think about this from maybe a North America perspective and also a global perspective, is the pipeline soft? Are we in trouble in terms of meeting demand? Then secondarily, there's historically been a significant issue with diversity in kind of technical and data in particular, are we getting better at that over time?
Danielle Crop (25:56):
I mean, first of all, I think that the talent issue is, I mean, pretty obvious, right? I think we definitely have a talent challenge within the areas of data, right? Whether or not you just think about that as the population of individuals who are choosing to go into these fields or you want to look at it from the diversity angle, either way, we have a challenge. I think that it's becoming increasingly more challenging with the fact that disciplines are getting more and more individualized, right? More specialized in nature. When we think about even within my organization, we have machine learning experts, AI experts, right? We also have the experimental design, all of these things came from very similar mathematical, statistical backgrounds. But they've all become disciplines in their own right.
I think the combination of the specialization of these different data disciplines and there's just the number of people who are going into them and the diverse, what type of backgrounds those people are coming from is definitely creating more of a talent crisis. I think that that's something that we and the universities need to start working on and I appreciate all the STEM work in lower levels and lower grades that's going on around the country and around the world to encourage women and minorities in STEM. I do think that will make a difference in the longer term, but in the short term it's definitely challenging.
Chris Knerr (27:34):
I mean, I think that a couple of really important observations just to start with the last one. I mean, I agree and in fact, this seems obvious, if we can improve diversity throughout the pipeline from early education, that's going to help to some extent alleviate, but it's kind of a long-term game. Unfortunately, my observation is, we've still got our ways to go in terms of the sort of winnowing out of diversity as you get to senior levels which is a chronic issue that's been talked about a lot.
I think your other observation is spot-on too, is that at consequent attendant to exponential growth in technologies, there has been a trend towards hyper specialization. This is actually in my view, occurred kind of [inaudible 00:28:21] if you will, the technology offices, regardless of how you construct those but then, because of the demand in a way is perhaps even more acute in the data fields which are almost just emerging like in the last 10 years as kind of a thing, as opposed to a more traditional program. This is not an easy question, but beyond what's being done as you discussed in sort of early secondary education there are obvious things we can do as data leaders and data professionals to kind of help with this.
Is there anything we could be doing for mid-career people who are sort of seeing the magic of this and maybe pulling them over? Interestingly like I'd shared with you, I didn't originally have a technical background at all. I came actually into digital and technology from almost purely a business track and then sort of learned on the job over time. There is hope for people but I wonder if just from a human capital management, there's more we could be doing kind of along the lines of drawing in perhaps non-traditional sources of talent to alleviate what is to your point, a pretty obvious pipeline shortage that's coming in the future?
Danielle Crop (29:45):
Yeah. That's a very interesting thought as well. I don't know that I ever thought about bringing people over mid career right into, but I think that's a very interesting idea. I think that the things that I've tried to do was more just be a representative of women in data, women in technology, women in digital, and then part of different groups both internally and [inaudible 00:30:12] externally, whether it be women in product or other things to make it clear that there are women in these fields and show other women that it's possible, right? I think if people see it, then they can do it, but if they don't see it, it feels impossible.
Chris Knerr (30:31):
Absolutely. Yeah. Absolutely. Hopefully over time, we'll be able to reduce kind of negative self-selection by having good role models and that's very much the hope. I think that's terrific. Maybe just to wrap up, can I put you on the hook for any major productions on say, a five to 10 year horizon for data science and analytics, either in financial services or in business and society at large?
Danielle Crop (31:02):
Well, I think that there's obvious one that is probably not a prediction that pretty much everyone in the field is, there's going to be more AI, right? This is going to become more a part of the bedrock of the technology that everyone is using. It already is. I just think that it's going to become more. I think that with that, brings more interesting questions about data ethics, right? I see that those things are going to continue to evolve. I also think that, we talked about specialization. I think there's probably going to be continued deeper specialization whether it be in data governance, data management, machine learning, experimentation. I think there's going to be continued specialization in the area of data. If there are any folks out there that are looking for right career change and listening to this, like for our conversation earlier, data is the great place to be.
Chris Knerr (32:06):
It's a great place to be. It occurred to me when you were answering, maybe to come back full circle. One of the things I love about this conversation, Danielle, is that we got into this sort of the central model and kind of the human touch. Is that something, if we look five, 10 years out as data leaders, we should be thinking about like ubiquitous AI, just from a science and technology standpoint is fascinating. On the other hand, I think that we can see the potential for businesses and society and tech companies to get carried away with that. I wonder if we would collectively take a position to say, "Look, here's one of the best companies, the best brands in the world, and a core part of their philosophy from the leadership down is that yes, all the automation, all the technology, but don't lose the human touch." I wonder if that's something that as leaders we can encourage in our fellow AI and data sciences as we move forward into a future where, to your point, there's just going to be more and more AI infused into everything that we do.
Danielle Crop (33:21):
Yeah. I think we have to always remember, right? That this is in the service of human beings, right? Not in service of the technology itself. I think it's easy to get lost in that, in the interesting problems that you can solve. But if it doesn't solve a human problem, then it probably doesn't need to happen right?
Chris Knerr (33:43):
Yes. I completely agree. Thank you for that. I think that's inspirational on a terrific point for us to wrap up today. Thank you so much for joining me, Danielle. Terrific conversation. I enjoyed it very much.
Danielle Crop (34:00):
Yeah. Thanks Chris.
Chris Knerr (34:02):
Yeah. For our audience, we've got some additional interviews at cdomagazine.tech. Again, thanks so much, Danielle, and thanks to the audience for joining us today.
About the AuthorFollow on Linkedin More Content by Chris Knerr