Perfume, Power, Prediction: Inside a Luxury Giant's Data and AI Strategy with Julie De Moyer, Chief Data Officer of LVMH Beauty

0:00:05.3 Satyen Sangani: Welcome back to Data Radicals. Today I'm joined by Julie De Moyer, Chief Data Officer at LVMH Beauty, the powerhouse behind 15 iconic beauty brands. Julie shares how she's leading data and AI innovation across the entire value chain, from product creation to customer delivery. We talk about how luxury brands blend human creativity with machine intelligence, how the CDO role is evolving, and what it really takes to make data a strategic advantage. If you've ever wondered how data shapes the future of beauty, this is an episode you don't want to miss. 0:00:40.8 Producer: This podcast is brought to you by Alation, a platform that delivers trusted data. AI creators know you can't have trusted AI without trusted data. Today, our customers use Alation to build game-changing AI solutions that streamline productivity and improve the customer experience. Learn more about Alation at A-L-A-T-I-O-N.com. 0:01:01.5 Satyen Sangani: Today on Data Radicals I'm thrilled to welcome Julie de Moyer, Chief Data Officer at LVMH Beauty. Julie leads data analytics and AI transformation across more than 15 beauty brands, helping enhance desirability, deepen customer relations, and elevate human creativity. 

Before LVMH, she had leadership position roles at Nike and across startups and global brands. She's a seasoned strategist with deep experience in data, customer experience, and digital transformation. Julie, welcome to Data Radicals. 0:01:32.0 Julie De Moyer: Thank you for having me.

Julie’s background: From economics student to technology leader

0:01:34.0 Satyen Sangani: So before we start, I'd love to hear about your career path because I think, especially as a Chief Data Officer, everybody wonders how did you get there, what took you there? And so tell us a little about yours. 0:01:45.4 Julie De Moyer: So the first thing to call out I am not your typical data officer with a fast track in computer science and then always been in the field. So my background is actually in economics, so I did a Bachelor and a Master's in Economics and later on an MBA. So I'm much more on the business side of things. 

But no, I started my career in product marketing in firms across the world and then later moved into the data, data and digital world. I've been working across from startups to public organizations to the FTSE 500 and so on. And lately I have been with some major brands in-house. So it's been a fantastic journey, an irregular one, where I have been able to work across domains across industries, but the red thread is that data and AI to just drive those business decisions. 0:02:33.0 Satyen Sangani: It's funny that you mentioned that you're trained in economics and not a computer scientist. I actually say the same thing about myself. I came to running a technology company and don't have any academic background in technology.

What is it about the economics background that had you gravitating towards data and, or, was it in spite of the economics background? What, how did you draw that line? 0:02:55.5 Julie De Moyer: I think where it started for me was it was more in high school, I took the classics and the, the maths track. So I had the maximum on maths. I was staying longer on the Thursdays in a Catholic girls' school in Belgium in order to just do more algorithms and so on. So I was really into that. And then later in uni, I started on my own sideway, I had a little running hobby here, doing some regressions on my own running data in order to run faster and to know the variables that could make me finish my marathons faster. So it's more of a hobby on the side. 

But it was only when I went actually to the US when I really got into the field, got into the startup scene. I followed my husband there back then. So it was back in, let's say like over a decade ago. And then I really got the passion and started to almost integrate that into my jobs. And then after a few years in the US working for a big public institution, the World Bank, where I was leveraging data and digital in order to achieve the SDGs. 0:03:54.2 Julie De Moyer: So very much around energy or around windmills or mini grids in the Kenya, it was really varied in order to just leverage that data to make those decisions. And then I came back to Europe at the same time to do the same thing for, in a consulting firm, where I was going into the commercial side and advising people on data and strategy around how to leverage the data in your organization to make those better business decisions. 

So to come back to your initial questions, the economist in me or the business mindset in me, it's really, really important because the classic 10, 20, 70 rule where you've got 10% is the tech, 20% is actually your data readiness, your processes. So it's, which is already the business and then you've got the 70% which is all about the people making sure that you change, you adopt, but that you also solve for the right question. And I think as an understander of the P&L and understander of the business needs, that comes in quite handy.

How has AI impacted the chief data officer role?

0:04:48.4 Satyen Sangani: Yeah, and how is that change management, sort of people understanding, informed how you do your work? Because, we're at a moment right now where we've had a whole bunch of chief data officers on the show, and many of them are talking in some cases about the demise of the chief data officer or how the role is under threat, and in many cases, a lot of the diagnosis for this situation is due to over-reliance or over-emphasis of technology, over-emphasis of sort of the tools, and under-emphasis of actual value. Talk about what you do and talk about how you think about this sort of transformative experience, and what's your process, and how does it differ from maybe some of your colleagues? 0:05:30.5 Julie De Moyer: Yeah, yeah. The first thing to say is that we've really moved that role as a chief data AI officer, and the term that is there, from being that builder to actually that value creator. So before we were more seen as like the classic, I would say, working on the classic projects, or the long-term projects, often very silent, very big, sponsored by the CFO, CIO, mainly on the process optimization. Very, very different from what we're seeing today, which is much more around the value, where you're working with the full C-suite, where I'm the trusted partner of the CEO, in order to achieve their three-year goals or their ambitions, which is a very different type of, I would say, mindset in that sense. And the leadership is also very different. It's no longer that I'm working just with the CIOs, but I'm actually really much closer to the business owners and the domain owners. 

I think there's also really a big cultural shift in how we then work as a CDO, where we started more from those siloed projects, big projects, long-term, often bit slow, or more like your typical ERP or your CRM systems that you would be integrating, less agile or less on speed. 0:06:35.2 Julie De Moyer: And where we're now seeing it's really about that cross-collaboration across your value chain, across your industries and your different domains, to much more speed and consumer centricity, like really listening to what is needed for the business, rather than the big tech implementations. And both are necessary, but the role of that data officer shifts more to that value and the impact to be shown. And I think it comes back to a CDO more as a sort of a value shaper, but also as an orchestrator, more of a real-time decision-maker. I mean, a really nice analogy I heard from one of my teammates, actually, in my teams, who said, like, yeah, we're no longer the rear-view mirror, but we're almost the augmented GPS. So we're much more going in the future and helping making those enhanced decisions. And I think we've also had a role to play compared to the past, where it was much more around big transformations and doing everything and trying to do everything. That's when we got excited about AI. Here, there is a big need for the CDOs, like myself and others, to also set those guidelines. There's a thing on democratizing and having all that chaos. 0:07:45.7 Julie De Moyer: But you got to get the guardrails and how far do we go with the AI? What are the ethics? How do you set that moral compass to your organization?

AI in action: How stakeholders collaborate

0:07:55.6 Satyen Sangani: Yeah. What does that look like? I mean, how do you partner with one of your business colleagues? And what does that interaction look like? Where does it start? And maybe how does the current state differ from how you used to work with these larger or big transformations, where it's all about sort of the project and not about the business initiative? 0:08:17.3 Julie De Moyer: Well, the classic, I mean, when I started, I think one of the turning points was really around, like, I think back when I was at Philips in the 2010s. So I had landed at Philips, I think, after a couple of jobs in Amsterdam, and I joined the product marketing department. And I was leading data for the product marketing department. So I had two jobs. The first job was manipulating the third-party data, so the Nielsen's of the world, to make sure that we've got the competitor intelligence and modeling on that in order to then have the right conversations with the wholesalers. And I had, on the other hand, I had this Six Sigma project, which was all around, and it's actually a statistical term, where you would go and model for, okay, how can we have the highest quality but the least variation? So almost about the product process optimization. So two really cool projects. And there that I actually saw a big difference between those leaders that were more top-down, more risk-averse, more controlling. My manager then was very different, was more democratizing, more that self-enablement. 0:09:13.2 Julie De Moyer: And I think that's where we're going to. That was, for me, the, okay, this is the way I want to lead. This is the way I want to move and work with my teams. And ever since, for the last decades, I've been working like that in a way of sort of a servant leadership towards your domain owners. You're sitting next to them. You're listening, mainly listening. A very, I would say, underrated skill to just understand what are these brand values? What do you want to keep? What do you want to optimize? Where do you don't go and where you stay out of the way? And then understand which decisions are the most valuable to actually tackle with data, AI analytics, and so on. 

So how we work is a very simple process. Is it always the same? Maybe not, but it's always going back to the same building blocks. First of all, sitting into those three-year plannings, listening to those CEOs, what are their biggest headaches? What are the line items on their P&L that they worry about the most? Or where do they see the value more in an intangible way where you can just hook on on a hook? And then it goes into a sort of a prioritization phase where you have many ideas and you obviously across domain need to see, what do we prioritize based on not just the impact but also on the complexity of some of these data models? 0:10:24.7 Julie De Moyer: You might have the data ready, not ready, it depends on how dependent you are, B2C versus wholesale, for example, in the case where I work now. And then we move into a sort of a build iteration phase where it's literally your data scientist next to your business domain lead in order to have and feed those algorithms with the right boundaries, the right parameters to make sure that you're still answering that right question. And on the end, it comes back to how do you then build that into the workflows? You're not just building an AI and then delivering your I would say your AI package but how do you then change that process along the way? 

So if you're working with marketeers, if you're working with let's say the producers or the perfumers, that will be very different. In the factories or in the stores, you wouldn't want the store manager to be on the three different apps in order to make the right decision or the right recommendation for a client. You want to build that in. And there again, it is a close collaboration between our data strategists and our data scientists to those people who are actually the users of the products. 0:11:26.3 Satyen Sangani: Yeah. I mean, that's a phenomenal process. I guess maybe let's even start with the very beginning. 

0:07:55.6 Satyen Sangani: When you partner with the CEOs and the CEOs are speaking with you, it takes the best analysts and it sounds like you're yourself in some ways an analyst because the best data analysts do, people will come to them with questions and ask about things and often the first question isn't actually the issue. The issue is sort of two or three layers deep. And then you have to sort of get at that point. And so what are the tricks that you use to do that? And what do you find are the best ways to get to the core of what actually matters? 0:08:28.4 Julie De Moyer: I think different ways of doing this, but what interests me the most, and I think my teams would say the same, is first listening to the process, describing it, mapping it out, and then testing it from different angles. So where you would have the department or the HQ are typically say in the process that runs like this, you would go into the geographies that could be in Shanghai, that could be in San Francisco, just to see how is that different in the regions. And you add that extra layer, and you peel that extra layer in order to see how is that different? Where are the process tweaks? 

Then you layer one more deeper where you would go into the stores. You're going to sit next to the maker. You're going to go away from the HQ or from the actual brand where they set the, I would say, the intended processes to go to see what is the real process actually in the stores. And often you would see that the intended and the real might differ a little bit more than you expect. So you would again layer that onto your processes.

How to optimize data-backed processes: Get out of the cubicle!

0:09:20.7 Satyen Sangani: Can you provide an example of that? When do you see the intended and the real differing? What's a good example where you've done that? 0:09:26.4 Julie De Moyer: Good example. In a recent company where I worked as sports, manufacturing goods, we would have in the HQ, they would think, okay, when we are delivering goods, we could potentially, we deliver them, they would go to a certain docking station, they would come down into the warehouse and they would be left on a certain place in a certain time and it would take an X amount of time to actually have that lead time through for how the warehouse in order to then be shipped for our delivery at home, for example. 

But when you actually go to the floor and you go to the port where all the goods are coming and you follow that package, those lead times and the hiccups might be very different than what you have on the paper. That could be from there is no load on a certain day, so you have an empty picking pool, which is, I would say, where you would need to go to the root cause on oh why is that the case, but it could also be that the machines are inefficient and that, let's say, the workers are not using the tools and go again to manual because they don't trust the tools that they have been given. 0:10:27.6 Julie De Moyer: There's often things in other companies that I’ve worked, it was more around where I would say they would have on the ground, on the working floor, found quick or ideas around systems that wouldn't work, so they would have fixed the broken process with kind of a tape on it, with that, I intended to say, let's say, one of our retailers where we would have, we have the cashier who normally need to scan and then they would have sort of expiration dates in order to see like how the inventory was still good to sell or to be sold or need to go out and where due to time issues or due to the machines not always working, they would have their own coding system and colors and dots around the store in order to make sure that we didn't sell anything that went over time and they would have actually find a plaster which wasn't necessarily digital instead of flagging it to the HQ that there was an issue. So it's often these minor things or these operational people on the ground that have good ideas that can then turn into a process optimization and vice versa that can happen on HQ as well, obviously. 0:11:35.5 Satyen Sangani: And many people in today's day and age deal entirely in the digital domain. And when people think about digital transformation, you're trying to move things to the digital world and you're describing something that's very physical. I mean, you're literally taking a process and then you're just saying, okay, well, this is a model of the world. I'm going to go actually walk this model in reality to understand step by step what's happening and speak to the individuals on the ground. 

I mean, do you enjoy that work more than... I guess that requires a lot of patience and curiosity, and even a little bit of empathy, and really just a detailed eye because many people will miss those details that, to your point, are the cause of something going off of the modeled path. And so, how do you do that work? And do you enjoy that work or do you partner with researchers to do that work? I mean... 0:12:25.7 Julie De Moyer: You partner with domain experts. Historically, data was set in your IT department and was coding and it's kind of delivering the IT projects and you would go through a ticketing system and you would just go work your way. Coders might do that at night, not necessarily be in touch with the brands. That was kind of how it was done in the past. That doesn't work anymore. We are here and in the brands that I work, they're not trying to put an AI on a broken process. They're really trying to go from end to end, from what are we trying to do to trying to solve it. If we use data AI along the way, that's great. If at some points the level of the process optimization versus AI is different, that's also okay. So we never know what we get ourselves into and we staff our teams based on what we see after the first, I would say, analysis. So I think what I get most to your question, what I guess more energy from is really going into that impact on the end. 0:13:21.7 Julie De Moyer: So how I would get to that is literally trying to first estimate that impact, making sure that we've got kind of work with finance and look at the volumes, what could we impact if we would look at a certain process, and then work with those domain owners, both on the HQ, when the stores or your e-commerce managers or your people in the warehouse or your perfumers or your R&D teams to go and fix an issue for them and then make that money and compare your estimations to your actuals. 

I mean, it's funny that you're asking what you're most interested in. Sometimes we have people on the team who enjoy being coders or being classic engineers who enjoys being in their cubicle, but we have more and more very outgoing and nice, I would say, generation of female data scientists, product owners, delivery managers, who more and more get the feeling, want to know about the brands, buy brands, especially beauty, it's skincare, fragrance, what I'm working in now, it's very relatable. And working with consumer goods, it brings them also a step closer to understanding where their AI makes the actual impact. So it's good for the person to feel meaningful, but it's equally good for the brands to stay connected to the real issue at stake, which isn't always the AI in the forefront, it's more the AI of the quiet AI on the back end.

The role of data in luxury today (and 5 ways to apply AI in retail)

0:14:42.0 Satyen Sangani: When you think about luxury brands, often you're thinking about art. I mean, there's a feel and a tactileness to the experience, and so much of the experience is in not just the buying experience, but sort of where it makes you feel and how it sort of represents, and the best brands, it is an experience. And you would think data would have a hard time proxying that, and yet it seems like data is actually so fundamental to the industry. Where does it matter, and how does data reflect sort of the art? Tell us a little bit about that interplay, because it does, in many cases, seem like it's coming out of some beautiful designer's hands into the store, and then you just get this experience. But there's a lot of science there, and so tell us about that. 0:15:25.7 Julie De Moyer: So, comparing different companies that I've worked for, right now, it's all around the craftsmanship, it's the quality of the product. The higher you go in the price point, but also in the luxury, I think it's really important to resonate with the consumers, to give something that is relevant. So, that would mean relevant qualitative experiences that last, that people want to remember, and that they feel comfortable with, that resonates with who they are, and what they want to stand for, what they want to project, but also what they don't want to do. So, it is very, very human-driven, and the AI is almost in the background playing a role, but not necessarily at the forefront. There's other companies that might see more the tech companies, that want to really shine with the AI, but that's what we're steering away from. That's just the vision that we have in our current company. I think the way you maintain that craftsmanship is, again, to my point, is setting those guardrails on where are you adding value, what does this brand stand for? How can you make the brand more unique versus making it more the same? 0:16:25.6 Julie De Moyer: Because you want to standardize, you want to find those synergies, but actually you do want to stay to that craftsmanship, to that really, the parfumeur or the vineyard driver or maybe it's the fashion maker that are actually building their soul into this and that are really bringing, I would say, craftsmanship into what they bring to the world, and then setting that against today, I would say the left and the right brain, that type of the brain, that creativity against those analytics, we are typically much more rational and much more calculated inside. So it's merging both of best worlds. We have multiple cases where that works. 

Along the value chain, I'll give you a couple of examples. If you look at the making of perfumes or the way we actually make the wines, in other industry, we would use the AI in order to help those, I would say, scientists to go faster, to optimize their trials. It will never replace the final sense or the final product that is decided on, but it can help with the substitutions of products that might need to go out as a result of regulatory changes. 0:17:28.7 Julie De Moyer: It might also help with making sure that the quality of the products lasts as long as possible. So we really help those research scientists do their job better and easier. But equally, when we are looking at the supply chain, it's the second area where we're working in, it's all around helping those planners to forecast better, to have less overstock, under-stock, to make sure that we're also inherently true to our values of being greener, being better for the environment. And that is one of the values of the company. So then our forecasts can help us achieve that. 

And if you, I mean, the way we talk about this is you've got the create, which is this R&D, you've got the make, which is your supply chain, then you've got the show, is basically how you show up. There you would have products that are helping us just allocate the rights, the message to the right person in the right place. So we really want to make sure that people are not overwhelmed or don't feel that we are contacting them without something that is relevant for them. So here we're really, it's really important in the luxury industry and beyond to be there when the consumer wants it and to also be aware or be okay with it that we don't message them if that is not what they are expecting. 0:18:41.4 Julie De Moyer: So that's the show area, the third one. And then the fourth area would be much more in the retail zone, or in the way we sell. That can be offline, online, different models, could be in a wholesaler or our own boutiques, it could be concessions, franchises. There's multiple ways of showing up and selling. And here we really want to make sure that that experience is memorable, that they can really experience the touch, the sensoriality of our products, and also understand that heritage of where does it come from, what is the story behind it. In many ways related to the art world where it's all around the maker, the intention, what do you do with it, when it's sold, who you're selling it for, the gifting, making sure that you have that experience not just for yourself but also for anybody you want to gift it to. 

And a fifth category where we use the AI that is then obviously in the continued engagement to continued relationships or your typical CRM or clients' connections. And that client connection can be nurtured by a variety of ways. The more you know about your clients, obviously never going outside of that privacy. 0:19:51.8 Julie De Moyer: It's always a give and take. As clients provide us with details on their routines or their ideas for an outing where they want to have a makeup look, we really use that information to their best knowledge in a secure way where we're providing them with the advice that they need in order to show up and to have a great night out or spend some time with friends and family. So it's all along the value chain and quite impressive. 0:20:18.3 Satyen Sangani: Yeah, and it's funny how much you mentioned not data but AI. I mean, three years ago, the job title would have either been Chief Data and Analytics Officer or Chief Data Officer. Today, it's Chief Data and AI Officer. And in this world, most of the transformation you described is primarily, the words you used were AI and not data. How much of your job is AI today versus data and how quickly has that shifted in the last two years? 0:20:48.0 Julie De Moyer: My job, I would say, it has different sides to it. I think there's a... Yes, we were more in the data world before, but I'm still very much in the data world. It's not because I use the word AI that I'm no longer making the data ready. 

So there's a bunch of time spent on making sure I've got the right data in the right format and that I'm combining data sets that are often sitting in siloed ways of the organization in order to bring that value that might be in threat or not looked at. So still doing a bunch of data. Equally, if brands are... If we are able to solve something with a simple if-then calculation or a simple analytics model or regression or whatever you want to do, that is fine for me. I'm not after the AI. It's not about the shining object syndrome. No, that is not what we're here to do. It's the thing that matters most and that helps the brand in a certain way. So I would say still doing a bunch of, I would say, alerts, triggers that already give the value. The AI is often the cherry on the cake. 0:21:48.5 Julie De Moyer: But more and more with obviously the arrival of GenAI and the agentic and so on, we're moving towards those new technologies that are becoming much more, I would say, available and helping us dream even bigger. So it's shifting, but not letting go of the data. That would be, as I say always, bad data plus good AI is still bad decisions.

Leading data in a multi-brand environment

0:22:09.6 Satyen Sangani: Yeah. In your case, just switching gears a little bit, you have an interesting dynamic where you have 15 brands that you support, and many CDOs, analytical leaders, have a tough enough time supporting one major consumer. In your case, you have to support 15. What are the dynamics there? How do you manage through them? How do you prioritize? Tell us a little bit about that work. 0:22:34.1 Julie De Moyer: Well, the good thing is not the first time I'm in a multi-brand environment. These matrix in the matrix organizations, a bunch of different styles from the times where it was brands like the Marces and the Heinekens of the world, where I would have an alcoholic beers and ciders, or I would have back in the Pepsi days, I would have like oats, and then I would need to sell Coke. It's very different audiences to tailor to. But it also brings a variety. It's almost, I left consulting with a bit of the, I want to focus more on my family and then the kids and be a little bit more at home. That was the reasoning. But I was at the time quite scared about the lack of diversity of the roles that I had to play. I was often in consulting, going on different transformations at the same time, and I could move at a faster pace with my teams. But actually, on the contrary, in multi-brand environments, the challenge is to find the synergies, because you're obviously wanting to leverage that one data lake or that one ethical guidelines or the one methodology to write your code, a central set of maybe data scientists and so on. 0:23:45.4 Julie De Moyer: But you've got to tailor it to those unique environments where you might have different types of data, different channel mixes, different product mixes in many cases, different price models, so the way we sell things, and then obviously also different volumes, quality. So the variables are endless, but it makes the job really, really interesting. It's like a never-ending puzzle where you cluster, I would say, the maisons at this time to not slow down the larger maisons who might have already the means and the bandwidth to go faster, but also to make sure that those smaller moving but very, I would say, agile, more less experienced but faster maisons can hop on the wagon of innovation. Yeah, it's an interesting question, but definitely one that keeps me also engaged in the role. 0:24:33.2 Satyen Sangani: Yeah, makes total sense. So you've now been a CDO for, and you are obviously in the world of... 0:24:40.6 Julie De Moyer: That might be a small exaggeration, Satyen.

How is the CDAIO role evolving?

0:24:43.3 Satyen Sangani: Well, and you are an experienced CDO, and so you're a CDAIO or whatever, the title probably is less important than the actual responsibilities, but how do you see the role evolving? There's a lot of talk about where the role goes, where it should be in the next three to five to ten years. Where do you see it moving, and how will it change, or how do you think it could change? 0:25:08.1 Julie De Moyer: So, in some companies, it sits in the tech department. So that would be literally still in the IT department where you're doing your job and you're already business-linked, like I mentioned, but you're kind of still tied to the IT role, and you would be reporting into IT and you'd be treated as a sort of a cost center. In other roles, I see CDOs that are sitting typically under your digital department. Because the way we've evolved, and mainly in consumer-facing companies, is from the CRM, where it was easiest to work into it, and then later spreading into the operational. 0:25:41.7 Julie De Moyer: So you would have a CDO that obviously is linked to digital and maybe less, I would say, into the supply chain or into the retail side of things. 

So where I see it moving is that CDO is no longer tied to certain departments as IT or being into digital, but there's actually a very much transversal role, which is already the case in the place where I am at, where you work across the value chain, as explained in all those projects, and where you're having a neutral stand, where you're just like any other domain, you're, I would say, a counter for the business value, the outcomes that you bring. You're the value shaper. You are part of the decision-making. You are part of the role shaping, and you're the one that actually brings that ecosystem together. 0:26:25.5 Julie De Moyer: So instead of sitting in a corner or in a certain department, you're moving much more to the trusted advisor of the CEO and the democratizer of the data across the department. 

And I think as a CDO, next to that trusted advisor, there's a big role to be played in the upskilling or in taking the company along on that journey. It's no longer AI projects from the top down, but there's a bunch of processes, a bunch of iterations, and the GenAI that makes it much more easier for actually your employees, the entry level or mid-level employees that can just come up with their own ways of leveraging AI for productivity. So you want to take those people along and make them understand that they themselves don't need to wait for the major project in order to optimize the way we work. That's where the CDIO, I would say, sits next to the CEO to understand, where are we going with talent? Who are we hiring? What are we looking for in people? And how can we make sure that the people that we have, have that basic knowledge on how to use the technology in a responsible way to do their jobs? 0:27:32.3 Satyen Sangani: It takes a pretty progressive CEO, though, to have a CDO that they would partner with them that way. And it also takes a pretty progressive CDO in order to be able to sort of have that business purview and have the understanding and the trust across the C-suite. I would say many have those skills. Many don't. What advice would you give to people who are sort of developing earlier in their career? How do you think of development? I mean, you mentioned that you work with a lot of female data scientists. And what would you tell them or others who are trying to get to the place that you've gotten to in your career? And I guess, how much of that is about sort of technology and capability? And how much of that is people? And how do you balance the two? 

How to become a trusted AI leader: Key tips

0:28:18.0 Julie De Moyer: Love this question. Three ways that you can maximize your career in order to get to any level, not necessarily the one I'm at. But if you want to grow up and you really want to talk about AI on a level that is C-suite and where you can just make an impact on the company. 

I think the first thing, it comes back to understand who you work for and where they make their money. So you got to really nail it there. Is it the consumer brand that values, say, the volumes that they sell? Is it a luxury brand that values the connections, that values the craftsmanship? Is it the packaging brand that values the speed? And what is it that your company values and how did that translate into a three-year plan on where they put their, I would say, their subject pillars? So the first thing is understand it. No matter at what level you are, understand what's the bread and butter of the company I'm working for and how can I be part of that translation? So that means saying no to things that are not within the pillars and saying yes to things that link closer to that, I would say, that's brand ambition. 0:29:15.6 Julie De Moyer: The second thing I would say is knowing that, what's the magic that we need to protect? Where are you going out of the way? Where are you going to go say no and don't do AI? And that really comes to, if you're in luxury and it is all around the quality, you will never choose speed over quality. If you're in a fast-paced, let's say an insurance company where it is all around making sure that people have the right insurance for their family model, you will never go too fast on selecting the insurance because you want to make sure that people have the right one. So there is different industries that maybe require different trade-offs. And once you understand those trade-offs, you'll be able to also be much more, I would say, eloquent in saying no to the things that you want to do. So understand the brands, understand the trade-offs. And then the third thing would be consumer first, always. So when you're working for a brand that serves a certain type of consumer, try to find some of those consumers in the region where it's more important, where the company probably makes the most money. 0:30:14.7 Julie De Moyer: Go speak to them, go understand what makes them tick, what ticks them off, and how you can add value to that consumer. And that's an iterative process because consumers evolve. They want to shop differently on different channels in different ways. They value different things. But if you can follow that as a third way, then you're going to obviously be successful in it. 

And maybe as a wrapper around that, it's great to understand the brand, to understand the trade-offs, to understand the consumer. You would want to be an invest in communication. That typically when you're coming from a data background, a technical background, you know almost too much and you become too operational in how you articulate yourself. So, how can you then step out of your comfort zone and articulate it in a way that is business savvy? So leveraging that brand knowledge, leveraging the language of your consumers, rather than talking in softwares or in Jira tickets or in algorithms or models. I personally, I don't use abbreviations anymore and I don't use softwares when I talk about AI. It's mainly about the impacts and the value that you're driving. And I've seen a big shift in how that is appreciated by the business by doing so. 0:31:21.3 Satyen Sangani: Yeah, it's super interesting because you described one of your early career experiences as a product marketer and it very much comes through in terms of how you articulate your advice because you're essentially saying, look, know the end value, know the differentiation, know sort of the core benefit statements that you're bringing out to somebody, understand that core customer and that user. I mean, all of that is, you speak in the language of a marketer, which I think is critical because I mean, if you actually talk to lots of successful chief data officers or people that are even not finding success, a lot of what they don't realize is that much of their job is marketing because so much of what they're trying to do is really understand the core need and then process that need to be able to then come up with a data product or something that is, people, we use the word data product now, but most of the time it's a report or a dashboard or table, but it's super interesting that you speak in that language and I think it's a great way to frame it and fabulous advice. 0:32:16.7 Julie De Moyer: And I think there's also some of the myths with data that I just feel like you got to have a lot of data in order to do something. Well, that's a big myth. You don't need the big data. You can just start with a qualitative set where it's just accessible, where you can really make a difference. So we don't need to wait for perfection. We can just try to do something already from the start. There's a paralysis of the imperfection, it's innate in anybody who is analytical or technical. We want to have the right, would say the confidence levels and so on. Get away from that and feel comfortable with following your guts and the brand guidelines, obviously, in order to test and trial. And I think as well that in addition to that, sometimes we feel we got to push data everywhere because our title is chief data and AI officer, but sometimes it's the data that guides and if there's no need for data, well, let's go with the gut feel. Let's go with the years of experience, some of these domains. It is okay that some of our planners still overwrite the ML sales forecast. 0:33:16.5 Julie De Moyer: It's okay to accept that and to go at the rhythm of the business. Yeah, it's a mindset shift. 0:33:24.0 Satyen Sangani: Yeah, and a fabulous one at that. This has been a really fun conversation and it's been pretty clear to understand why you are where you are and doing what you're doing. So Julie, thank you for taking the time to speak with us. 0:33:34.4 Julie De Moyer: Likewise. 0:33:36.6 Satyen Sangani: Until next time. 0:33:38.1 Julie De Moyer: Yes, see you next time. 0:33:40.6 Satyen Sangani: That was a fascinating chat with Julie. She reminded us that behind every great AI strategy is strong data, clear ethics, and above all, human creativity. Julie also stressed the value of getting into the field, talking to real people, and uncovering the gaps that data alone can't show. In the end, it's all about keeping the customer at the center, respecting their privacy, listening to their feedback, and constantly improving. I'm Satyen Sangani, CEO of Alation. Thanks for tuning in to Data Radicals. Stay curious, stay bold, see you next time. 0:34:14.7 Producer: This podcast is brought to you by Alation. Your boss may be AI ready, but is your data? Learn how to prepare your data for a range of AI use cases. This white paper will show you how to build an AI success strategy and avoid common pitfalls. Visit alation.com/AI-ready. That's alation.com/AI-dash ready.