Stay up to date on the digital shelf.



    We'll keep you up to date!


    A Model for Global Analytics Transformation and Harmonization, with Celia Van Wickel, Director Omnichannel Digital Commerce Analytics at Mars

    For a company that has been around since 1911, Mars has a reputation as one of the most visionary companies in the industry, particularly when it comes to building global capabilities that scale and flex to serve the needs of the regional lines of business. It’s not any easy feat. Celia Van Wickel, Director Omnichannel Digital Commerce Analytics at Mars, drives building out their global digital commerce analytics strategy, and works closely with the business to design, test and scale capabilities that deliver efficiency, measure results, and identify opportunities for growth. She kindly agreed to join the podcast to share the art and science for enabling digital commerce analytics to every region around the world. 


    Our transcripts are generated by AI. Please excuse any typos and if you have any specific questions please email info@digitalshelfinstitute.org.


    Peter Crosby (00:00):

    Welcome to unpacking the Digital Shelf where we explore brand manufacturing in the digital age.


    Hey everyone. Peter Crosby here from the Digital Shelf Institute. For a company that has been around since 1911. Mars has a reputation as one of the most visionary companies in the industry, particularly when it comes to building global capabilities that scale and flex to serve the needs of the regional lines of business. It's not an easy feat. Celia Van Wickel, Director, omnichannel Digital Commerce Analytics at Mars drives building out their global digital commerce analytics strategy and works closely with the business to design, test and scale capabilities that deliver efficiency, measure results, and identifies opportunities for growth. She kindly agreed to join Lauren Livak Gilbert and me to share the art and science for enabling digital commerce analytics to every region around the world. Celia, welcome to the podcast. We are so delighted to have you here. Thank you so much.

    Celia Van Wickel (01:11):

    Thank you for having me. It's a pleasure. Big fan of the Digital Shelf podcast. Oh gosh, really excited to be here and to give whatever insight I can back to your audience.

    Peter Crosby (01:23):

    Well, you are doing some incredible work at Mars around data and analytics, and as we talk about a lot, brands have more data than they can use, but they're really not always set up to layer all those data sets together and really form a clear picture of what's happening in the business. And pulling that off I would imagine takes a lot of collaboration internally and then a lot of collaboration sort of create and get the impact out of it. So you have built this analytics team, and so we would love to know more about your team and how they serve the business.

    Celia Van Wickel (01:59):

    So specifically at Mars, I lead a global team that focuses on digital commerce and omnichannel, advanced analytics activation, particularly in the Mars snacking division of Mars. So I'm actually part of a larger group collectively where our goal is to drive one demand and data and analytics across all functions, what you would call the demand side of the business. So sales, commercial marketing portfolio. We have various folks who are supporting the one demand side and together working collectively as I think about my digital commerce agenda and my team's agenda to drive the strategic goals. So I personally guide my direct team to think about strategic digital commerce data and analytics solutions and vision and how we build best in class solutions towards these one demand goals. So we're enabling digital commerce by building business intelligence tools, analytics solutions to really give data at the hands and fingertips of sales leaders for Amazon omnichannel leaders and Sam's Club people leading Instacart and Uber Eats and other on-demand delivery platforms globally. And we're also working very closely, very close to your heart in the digital capabilities area around search content management and really building our own best in class solutions for digital shelf measurement, what we call here at Mars Perfect digital store capabilities.

    Lauren Livak Gilbert (03:25):

    So Celia, a question, when you think about having a global role, does that mean say I am a digital lead in the business on one of the brands and I want to know how to connect sales to my content on our PDP? Would they come to you and present that problem and then you would work with them? What is the dynamic between the global and the regional activation of the data?

    Celia Van Wickel (03:49):

    So each of my people and my team actually serve a different region of the world, but I actually serve globally up into a cohesive strategy. So first we start with what is the global strategy we're trying to achieve? What are our biggest bet markets, maybe our growth in activation markets, and kind of come up with a plan there in terms of how we're going to globally roll out different solutions. Sometimes we'll have markets also come to us and say there's a need and we always want to try to unify it back up to that global strategy of what we're trying to achieve and how we are getting there as a group. I've been on a journey, so the group I joined was brand new when I joined in 2022. So we've been on this journey of how do we enable the regions, especially the US, Europe and some other regions of the world now, how do we really think about that global unification and strategy to harmonize and scale across the world from the biggest markets to some of the smaller markets, whether that's in Asia or that's in Europe or other parts of the world.


    So we're really looking at that strategy today, but again, it starts with global strategy first. What are we trying to achieve? How did the markets and regions enable that? And then we kind of work with those regional partners to help us activate and buy into the solutions so they can actually build the next best in class measurement for their areas as well.

    Lauren Livak Gilbert (05:04):

    I think that's so unique. I mean from a global analytics function, especially if you're thinking in the digital commerce capability space, because I know I work with a lot of brands who say they have to figure it out in their team and they don't necessarily have that global support. And I can imagine you can then provide a lot of perspective around organizational data like sales or SKU count or prioritization of SKUs, and that can create a lot of, or it can take a lot of the responsibility off the brand and then you can provide that to them while they're doing their day job. Was that kind of the premise? Take it out of the region so they can focus on activating and you can have that broader umbrella.

    Celia Van Wickel (05:48):

    So I think it's about more solidifying unified strategy, but the regions still want to be enabled. So it's not to really alleviate them necessarily, they want the enablement, it's more to be more cohesive and purposeful about what we're measuring and how we're rolling that out across the world. So we're all looking at our performance in the same way. So you talk about it a little bit further about how we're doing that in the digital shelf, perfect digital store space specifically. But in this area, whether it's sales data or it's digital shelf data, every market and region has different data depth quality or sources that they're using for. So if we can unify what we're measuring globally, what we're trying to see and unify globally, then we can bring that down and use the data that they have available to enable that solution. And if they can't enable that solution we come up with, we had to work with a business to come up with a plan, like a data plan to achieve that.

    Lauren Livak Gilbert (06:50):

    You're all marching towards the same thing. We just talked to someone on a podcast the other day who said there was a different definition of, I think it was net revenue across the business, which makes it really challenging to define what that means and then to look at the same statistics across the board. So in terms of measuring, can you tell us what are some of the key areas that you're focused on measuring?

    Celia Van Wickel (07:12):

    So as I mentioned, we're focused a lot on digital capabilities is one core area. And so in that area we are definitely looking at unifying, at searching content drivers across the different markets. We're starting also to introduce things like anomaly detection for content, making sure that PIM Salsify type of integration is there from a gold standards benchmarking perspective globally where we have it. And really again talked about earlier, unifying those KPIs. So you mentioned the net revenue for another company kind of being different, right? Well, we have share of search being different everywhere or total health score being different everywhere because every business has different perspectives or every data source and provider has a different perspective on how they measure that. We also can look at the raw data and assess quality issues and harmonize that data as well. So we can pick up if there's a skew off or something's missing from a data.


    So it really helps us understand and apply those weights equally, those KPIs equally, and really focus on, okay, what's the best in class, what's the right unified north star, what we should be achieving from a KPI perspective? And then we execute on those performance scores across each market. So when we're doing what we're working on is that US to Europe, to Mexico, to South Korea, we'll be measuring content search the same regardless of whatever data source they have within their ecosystem. And so that's one of the core areas there. And then where we go a little bit deeper is we have different tools. One is called search algorithm decoding. And so we're really actually going deep on different retailers search algorithms to understand the value drivers of that retailer. So whether it's car four in Europe, we can actually see the drivers from Uncoding that search algorithm and we'll know whether it's our content presence or it's our sales drivers or it's other types of taxonomy that we need to drive within that retailer ecosystem.


    And then another one that's always fun is that we also measure incrementality of our onsite paid media measurement. And so we're actually looking deep by tactic by keyword for overall incrementality performance. So we can see that for the Tesco, Amazon and others to really help the e-commerce teams along with their shopper marketing teams really understand how to optimize that mix to drive onsite sales performance. And so again, there's a nuance there because IT team focuses on digital commerce. So we're really focused on the customer's online sales perspective in terms of what we're measuring and to show that case of incrementality. But incrementality could actually be shown across different use cases depending on what you're trying to actually measure. And lastly, the other thing is I also enable sales teams. And so one of the things we're doing with sales teams that actually could connect back to our digital shelf and perfect digital store measurement is that we're actually connecting solutions end to end to understand those drivers.


    But to do that, we're first trying to drive and understand the e-commerce p and ls for every customer from almost around the world. So those have been very hard to obtain to understand. You can understand holistic p and l, but to understand the p and l for e-commerce has been very disparate based on the data sources available. And so we're actually working a lot with harmonizing, bringing in our different e-commerce POS data and connecting those value drivers and those p and l drivers, which will allow us to then go back and really drive true understanding holistically end-to-end sales p and l of different drivers such as perfect digital store or portfolio or other category drivers that maybe the customer is trying to achieve. So that's a really big undertaking that my team is doing right now is really trying to unlock those key value drivers so we can understand the true health of a performance from a global perspective.

    Peter Crosby (10:55):

    That's really amazing. It is an amazing effort because part of what we've been, we talk about a lot here on the podcast is in this era of money's no longer free and digital commerce and e-commerce needs to mature and become part of an overall omnichannel profitability. I would imagine, and you can confirm this or not, but that these business shifts, which is a drive towards the whole business needs to mature and everything needs to come together into ultimately, and I don't want to sort of misstate your naming, but sort of a perfect store for the consumer no matter where they are and no matter where their journey takes them. And I would imagine these analytics one are incredibly difficult to get together and normalize and pull assets insights out of, but that the organization must be grateful to start seeing that really show up. I was just wondering if you can dig into a little bit about how this data comes to life, particularly the p and l data to help drive towards that profitability that you're looking at.

    Celia Van Wickel (12:04):

    So it could get very complicated,

    Peter Crosby (12:07):

    I imagine. So

    Celia Van Wickel (12:09):

    It depends also again on the data maturity of every market, what's available, but we start with the PO os data. We need skew level data to see that we need to understand how it maps to our cases, but then we have our own calculations for p and l and the KPIs and then we make sure that we're mapping that accordingly. But we need unified measurement, we need unified mapping. Sometimes we actually use our PIM data to map those SKUs. It's a very complex, it's actually a lot of things are in there. In fact, some of what we're finding is that some of the e-commerce po OS data is very volatile. So it's accurate one month and the next month it's not accurate anymore because the retailer may be threw in a SKU or two that's not ours. Our data feed system may not be capturing every statement that the retailer did.


    We are constantly troubleshooting this space. And so it takes a lot of effort and a lot of mindset to really drive that clean data for the organization to even get to the p and l level. So it's very, very complex. And you mentioned something earlier with one perfect store. And so again, my area is more in the digital commerce side. Making for how we show up online is perfect, but we are having active conversations. Going back to that one demand point of view is how do we look at perfect execution in store to perfect execution online? How do we look at similar but not exactly the same measurement and value drivers. So someone at the Tesco or Walmart or wherever can see side by side holistically what is going on from how we show up both online and offline. Again, very different how you measure that in the offline space and the in-store space, but share of the store to share of search, these are things of proxies that you can come up with and look like for in terms of total execution and value of business.

    Peter Crosby (14:12):

    I love the idea of reverse engineering search algorithms to, you may not put it this way, but I was thinking that sounds like a competitive advantage if you can figure out, and particularly you were talking about how much some of the data shifts and algorithms on, especially on the big players change constantly. And if you're able to, that just sounds super interesting. And I could imagine if I were one of those leaders who was able to have that data, I can adjust my strategy, my content strategy, my ad strategy. A lot of that sort of feeds off of, I imagine the signals that you get from that decoding.

    Celia Van Wickel (14:53):

    So no one's as fast as Amazon yet in the algorithms, but TikTok I think is coming, that's like an algorithm that it's almost impossible to unlock, but the retailers are changing them fairly regularly and we've done a lot of work deeply in Europe in this area, and we find that just looking at it at a certain periodic level allows us to kind of adjust that strategy. No organization that I've seen yet, whether I talk from my own or from any other organization, have I seen that you make daily decisions, you can actually actually take a daily action on your own, right? Everyone's very busy, but these algorithms, they change maybe every quarter to annual and we kind of take a look at them to make sure that we're always driving the right value drivers for that retailer and trusting their strategy

    Peter Crosby (15:44):

    That has AI written all over it. But that's another episode. Very interesting. And so I can imagine looking at this whole process end to end and all the teams that you're collaborating with and sourcing data and then getting it out into the hands of the people who need it is really a great example of building cross-functional and cross potentially sometimes silo collaboration within an organization. And how do you keep that collaboration alive? How do you make that work effectively to be able to impact the business the way you do?

    Celia Van Wickel (16:24):

    Yeah, so one of the things in analytics we really say is best practice is really understand the business needs first. So if I'm building perfect store or sales movement, what matters to that person? What are they trying to drive? How do we leverage interviewing, process design thinking skills to understand those business needs first before we go build anything? So that's really one of the things we do. But to even get to building something, once we understand the needs, it has to actually have people who want to take ownership of analytics solutions in the business. So whether it's global leading the strategy, but sometimes global again helps us set the strategy, but we still need those regional market level partners around the world to help us embed that connectivity of the data for their markets and partners. And so we need someone who has strong sponsorship and ownership of wanting these capabilities within their market and really helping us drive what we call change management training and enablement of these solutions once we build them to make sure that they're getting used.


    So again, we really want to deliver that, but we also need those top sponsors across each region bought into our harmonized solutions and strategy. So we're all driving towards the same thing towards the best in class markets. We might build best in class design in Europe first because they have more needs and some questions, and then we figure out how to scale that solution accordingly. So again, it is a constant collaboration about how we bring in the global to the regional and guide that true ownership in anything we do, right? If we can't get the market on board with the right people, we can't build it, we won't build it probably work relatively. We all have to be on the same page.

    Peter Crosby (18:19):

    That's a common theme. I mean, leadership is always a common theme right across these things. And finding the human beings who are engaged by the possibility of what could happen and then seeing their way clear to committing the resources of doing something like that, which might take a little bit to pay off or something instant among all of the work that they have to do day to day, they make a decision to take this on because they see what it could bring. And I was wondering, do you have a sense of the characteristics of those leaders that you've built, this core of leaders that are willing to be sponsors with you and go on these sometimes frustrating and difficult journeys to achieve these things. What are the characteristics of those leaders? What do you look for? What have you found in these people that help you with these kinds of transformations?

    Celia Van Wickel (19:14):

    Yeah, absolutely. So I would find that, I know a lot of people are on a journey being more digital data savvy within their organizations. There's something who kind of just lead that first, right? You always have someone who's more passionate about it. Those are usually key partners to start incubating with overall. But it really goes back to that buy-in of what the analytics solutions will do for their team. We need them to be embedded into business objectives. And so we kind of look at multiple layers and factors. One of them is whether we're going to drive time savings back into the organization. The other is business value. Will there be net dollars on the table for having these solutions? So when they can understand those kind of drivers, they then get more bought in for their team. I actually think time savings has more value sometimes than the actual value of dollars in the business.


    Yes, we want to drive dollars and we will drive dollars, but we need that time from people to give them back time within the solutions that we build for them. So we need those strong partners. We also have strong partnerships and people who can understand the holistic end-to-end view of how measurement will drive value to the business. We need people who can help us partner on roadmaps to make sure that we're driving it right for the business. So it is not necessarily an e-commerce person only, but we need someone who has that more forward thinking vision who might already be a good data enabler within their organization of what they're doing day to day to help us build that holistic strategy.

    Lauren Livak Gilbert (20:55):

    And how do you see that the regions report out on the data that you provide? Because I know that it can be challenging when you're sitting in the region to share out a scorecard and nobody knows the cadence to share that out and don't know who to share it out to. And there's fear that if you share it out and the number is wrong, then someone's going to ask too many questions and you can't provide the answers to that. So I'm curious how the regions and how you as the global team help to disseminate that into the organization in a productive action focused way. Not a, oh no, my numbers are red and we did this wrong kind of way.

    Celia Van Wickel (21:35):

    So it comes back to how we build it with a sponsor, a strong product owner who's going to help us do that. So they be, we help them understand what's in there, what the data is telling them when we build the tools and capabilities, but then we have to actually go out with them on training the teams. So they actually become the central point of knowing everything that's in that tool, knowing what it's measuring within the tool. And so we go back to their business enabled business owner, right, to help us with that. It does take some change management on their end to get people to use it day-to-day. We will help create some thought starters in ways to use those opportunities from the data that we're seeing. We're showing that value and lift. There was a tool actually that there was a question that came into the business about value of investment for one of our retailers.


    And we had built a forecasting tool, and it was used actually for different use cases, but playing around with the tool, I actually was able to answer that question for a different use case. And so how do we help partner with them? But it is a holistic training. We have serious training that we just rolled out to how do you be a business product owner with us in analytics and what's your roles and responsibilities for what you need to do? In some of our markets, we have full change management plans that business owner with us have to be bought into in terms of who's going to embed it, who's going to own the strategy with us to get it into the business. So again, we do look at that. We're also looking at usage of our tools. So we're tracking every 90 days or so, how much usage are we getting? It's the right users who are using it. We're doing adoption surveys to understand how our tools are perceived, how do we make them better? So we're doing these different things to get people in there, but thinking about some of the challenges that we have, it's not, I like this line from the movie Heal the Dreams with Kevin Costner. So they say if you build it, but from my case, they will come. And that's not ever the case,

    Lauren Livak Gilbert (23:40):

    Not with data.

    Celia Van Wickel (23:42):

    They need to be along with you to make them come. And so again, it comes back to that strong partnership to drive and embed change management and get those solutions bought in. So for example, our perfect digital store, we have a great person globally helping us think about the harmonization strategy. And we have a really strong capability leads who are very passionate about measuring their digital shelf KPIs and making sure that their teams are accountable for that measurement. So it's truly a partnership. I cannot do it alone and I will not, if I build it, it just won't be there to come for people to come.

    Lauren Livak Gilbert (24:19):

    And a huge kudos to you and your team, Celia. I mean, I think this is a very, very unique approach, the way your team is built and how you're working with the regions and the amount of data you're looking at at such a high level that you're able to answer these very specific questions that will inform the overall strategy of the brand. So I really think it's a very unique team. I haven't heard a lot of teams like this in the organization. And since we have you, can we pick your brain to say if you're in an organization that might not have a global analytics team or you're building a global analytics team, what are your lessons learned advice call outs to say as you're doing this, think about these things or this is what I wish I would've done when we started. Love to get your thoughts.

    Celia Van Wickel (25:05):

    Absolutely. So one, it's going to take time. It takes significant time, not only from my team, but it takes significant time from the business to help us build and embed these analytics. So people need to be bought in that they're going to take sometimes 30% or more of their time. I mean, it's pretty significant to help us build it, right? But I talk about is that value driver back to them, right? Does it save them time? So look at those hours back once we're done, or is that value that's going to drive by? Are we locking dollars for the business or other KPIs? So that's one of the first things. The other thing is that we have to build in incremental steps. So do we have the data foundation set to even measure the basics to bring the data together? People may not need right away the fancy AI model to be applied.


    They might just need that triangulation of data. So our internal data to the second party and third party data sources. And that has immediate unlock and value in itself. And then once we get that foundation set, we can start building the fancier forecasting tools and integrated driver tools and sales drivers. The other thing is that we really need to listen to the business to unlock value and understand strategy. So we need those people to help us understand, as I talked about with that retailer example, right? It was a random question, but we had the tools to do it and then that tool's being used or reconsidered for other purposes. And one of the other things that, again, each market, especially as we've done our perfect digital store work for global, every market has different data sources. So again, what are those unifying KPIs? Can we measure everybody in the same way?


    So again, we're looking at can we use not only the same KPIs, but what's the weight, what's included in that measurement? How does that come together? How does that globally scale? Maybe smaller markets just need a scorecard view, and that's all we're going to be able to enable for them. And as others, we can go deeper into the data for maybe some of the larger medium sized markets. So we're constantly battling this, but I think the thing is to think about things in chunks and bits and pieces to give time line data does not always work the way you want it to when you first start out. So again, do a data assessment. What is it we need to have in the data? Do we have the right data? And then what time will it take us to get to the data in the right place that we need to enable the business? So these are the things that we are learning as a team, and they have had challenges for us, but we also have learned to overcome a lot of them. We're very big on that. You learn from failure, you learn from mistakes. And so those are the things that we're trying to drive towards to drive this best in class solutions for the future and really do some cool things that we have in our North star roadmaps that we're trying to do in the future.

    Lauren Livak Gilbert (28:04):

    I love that point about the data. I was just going to say, because I think people underestimate finding the data. Where does it live? Who owns it? You can build whatever fancy model you want, but if you don't know where it lives in your organization and you don't have the right data governance, you don't have the right data dictionary, you don't know what you're defining with that data, you can't then take this effort that you're talking about Celia and make it a global capability. And I feel like that chunk of time you were talking about, a lot of that falls into just finding it and knowing where it is in the organization before even building this broader model. Would you agree that's the upfront time suck?

    Celia Van Wickel (28:44):

    It has been a huge time suck. Like I said, we're building these p and ls that I talked about, getting the data in the right way in the right format and then not having it change on us is a big one that we're trying to figure out and tackle and making sure it's a more fluid process. It is a lot of enterprise learning that we're doing as we're bringing this information on for the business. And so yeah, there are those challenges that we see there. Again, there's challenges in the measurement. There's different things that we have to decide and align on as an organization.

    Peter Crosby (29:17):

    Well, I mean for a company that was founded in 1911, I am consistently impressed working with Mars, their long-term commitment to just building and testing and expanding. And your organization, as I've experienced it, has a tremendous amount of ambition and vision, but also patience at the same time to take projects like these through and they end up really impacting the overall business in such a fundamental way. And I wonder if from the inside, does it feel that way to you? I mean, it feels like you have a ton of support from the organization to do this. I'm sure there are all the frustrations and wish I had more budget and more people and all that, but I was just wondering if it feels on the inside what it feels like from the experience of having worked with you all.

    Celia Van Wickel (30:18):

    Yeah, absolutely. Right. We're really here to drive transformational analytics solutions. They want us to globally scale capabilities to unify measurement, to have a holistic view of what's going on around the world. So they are supporting this, right? This is core and fundamental to the organization.

    Peter Crosby (30:37):

    Well, Celia, so grateful that you take the time to share this with a broader community. It's really generous and helpful, and we are really grateful for you coming on the podcast.

    Celia Van Wickel (30:50):

    Well, thank you for having me. This has been such a pleasure. Again, I really enjoy being here and I really hope that your audience learns something today. So thank you so much.

    Peter Crosby (31:00):

    Thanks Cecilia for sharing Mars approach with us. There's always more information on digital commerce excellence to be had at digitalshelfinstitute.org. Swing by and become a member. Thanks for being part of our community.