For a retail powerhouse like PepsiCo, which collects data from countless sources, the ability to transform complex data into strategic decisions has never been more vital. This is key to gaining a competitive edge, meeting market demands, delighting customers, and achieving a multitude of other business objectives.
Ellen Webb, Vice President of Shopper Analytics & Insights at PepsiCo, leads a team that turns complex data into actionable insights, driving strategic decisions to enhance both in-store and online shopping experiences. With her leadership, PepsiCo has gained a competitive edge in an ever-evolving market.
Webb’s strategy combines advanced analytics and machine learning to drive informed decision-making, seamlessly integrating diverse data sources—from proprietary household data to third-party insights. By fostering close collaboration with retail partners, her team ensures that shopper insights lead to meaningful actions across PepsiCo’s portfolio.
In the first of this interview with Todd Foley, Chief Digital Officer and CISO at Lydonia, Webb shares her unique perspective on how data and creativity are driving PepsiCo’s success and leadership in retail analytics.
Edited Excerpts:
Q. What led you to a position where you were taking complex data and applying it in such an immediate way?
A. I’ve always been drawn to roles that balance left-brain and right-brain work. I love diving deep into data and analytics — it’s something I’ve been passionate about. My curiosity drives me to explore the numbers, but I also thrive on the creative side — bringing insights to life through human understanding and strategic thinking. It’s about moving someone—whether an internal or external customer—to take meaningful action.
Functionally, my career has spanned both customer-facing roles, like category management early on, and internal roles, such as sales strategy, business planning, and, most recently, insights and analytics.
Channel-wise, I’ve worked across both retail and away-from-home spaces. The dynamics of shopping in a store versus at an airport, hotel, or business cafeteria are entirely different, and understanding how shoppers engage in these environments has been fascinating.
Most recently, I’ve added digital commerce and space transformation to my experience, rounding out an omnichannel, end-to-end perspective.
Our insights and analytics team sits within the commercial organization. Being in the room where problems are discussed helps us become better at what we do — solving issues efficiently and driving speed-to-action like never before.
Q. You’re focused on transforming retail spaces—not just deciding what to sell and how to sell it, but also how to present it in-store. Is that accurate?
A. It’s about taking the strategy and insights and translating them into effective in-store execution. When shoppers are standing at the shelf, whether in-store or online, we want to ensure they have the options they’re looking for.
A big part of our work is focused on meeting shoppers where they are, providing what they need, and supporting our retailer partners in the process.
Q. How do you combine traditional brick-and-mortar strategies with digital engagement, especially when using advanced analytics and machine learning for things like product placement?
A. It starts with the data. The first thing that we always make sure is adhering to our data privacy and responsible use of information policies.
It’s a privilege and not a right to have all this data. We take the utmost responsibility to make sure that we’re using it in the right ways. Speaking of data sources, we have a wide range, including open-source data, first-, second-, and third-party data, along with proprietary household data. We also have a ton of data that we have that gets created through our DSD or direct store delivery network.
The work that has to go into connecting those disparate data sources and making them into something valuable, is incredibly labor intensive. That’s where machine learning comes in.
The data has to be in a place in a way that is connected to do something with it. That’s where we’ve prioritized that work so that our teams are not spending their time and energy around data engineering. We’ve got a team that helps us do that because it is critically important.
My team then is focused on what we are going to do with that data. We’re very mindful in determining what are the data sets that we need to be able to answer our questions. There’s been data sets that we’ve said no to because they’re not to our business.
Q. How do you and your team work together to ensure they balance the technical aspects, like machine learning or data preparation, with your current goals?
A. It all starts with identifying the problem we’re trying to solve. For us, that’s deeply rooted in our Pet Biz approach, which focuses on partnering with vets and retail partners in a way that’s both insight-driven and performance-focused.
Our retail partners know their shoppers better than we ever could, but we bring an outside perspective. For instance, retailers understand how their shoppers behave within their four walls—whether physical or virtual. What we can do is complement that by shedding light on how those shoppers behave outside their stores.
This process begins with gaining clarity on the problem we’re solving. It helps us stay purposeful as we navigate the data and develop solutions that drive granular growth for our customers, working alongside them.
Where my team sits in the organization is unique—we’re directly engaging with customers, actively listening to uncover not just the explicit questions but also the unspoken needs. Sometimes it’s a matter of saying, “I heard you mention this—can we explore it further?” This discovery phase is crucial before we even start the work.
Interestingly, the solution isn’t always in the analytics. It could involve primary research or even third-party data. It all comes down to selecting the right tool from our toolkit to address the problem effectively.
Ultimately, it’s about partnership. In many cases, the process is iterative—we uncover something, present it, and then refine based on feedback. This collaborative approach has been key to how we operate with our customer partners and deliver meaningful insights.
Q. You seem to be describing a very mature data and governance environment where agility is key. How would you assess your environment on the maturity scale?
A. It’s always possible to do more. We’re fortunate to have excellent partners in our strategy and transformation team who take on much of the governance work. That’s not something I manage, but we benefit tremendously from their efforts.
Our focus is more on the data itself and the solutions it drives. For instance, I’ve mentioned before how we aim to scale these solutions across the organization. A significant part of our work involves organizing our analytics solutions around core business problems.
For example, are we looking to drive more trips or build larger baskets? Identifying these high-level business questions helps us guide people toward two or three key solutions as starting points. This approach helps spark conversations, especially with those less familiar with the analytics side.
It’s very much an iterative process. We might begin by considering one way to address a problem, and then refine it based on feedback or changing conditions. Agility is key—understanding the problem, the surrounding context, and the best way to narrow down and tackle it effectively.
Q. I’ve heard you mention the term “phygital” — could you explain what it means?
A. Phygital is essentially about blending physical and digital spaces. It’s not an entirely new concept—it’s not groundbreaking in that sense—but it’s also not something far off or futuristic. It’s happening now.
For us as an industry, it’s about rethinking the digital experience strategically. This involves leveraging digital systems and data to enhance not just online commerce, but also the in-store and pickup experiences. The goal is to make every shopping touchpoint—whether online or in-store—a seamless and intuitive experience for shoppers. It’s about bringing what you see on the shelf in-store together with the digital world, reducing friction and minimizing frustration for customers.
Q. With the vast amount of data generated today—from every action, whether online or in-store—how do you guide your teams to effectively mine and use this huge volume of information?
A. My team’s primary focus is to advocate for the shopper—that’s at the heart of everything we do. When we talk about digital shoppers, they now make up the majority. In fact, 63% of shoppers are grocery shopping both online and in-store, which represents a growth of about two and a half times since 2020.
This behavior has become nearly universal and will only continue to expand. At this point, it’s not optional. Ultimately, if people have a good experience, they’ll reward you with their loyalty. That’s why ensuring a seamlessly blended online and in-store experience is so critical.
A key part of achieving this lies in understanding the data streams. What do we know about the online shopping experience compared to how people behave in stores? By analyzing this, we can make sure that we’re making logical leaps within that space.
Q. Given how much the market has shifted, especially after COVID, with changes in where people live and how often they go to the office, it must also influence shopping behavior. How does your team track and project these changes?
A. We do a lot of that. I think one important thing to remember is the statistic I mentioned—63 percent of shoppers are engaging in both online and in-store shopping. In the insights world, it’s crucial to move beyond what we call a “sample size of one.” That means stepping away from personal experiences or assumptions and focusing on what the data is actually telling us.
Take Gen Z, for example. While they obviously enjoy online shopping, they also value the in-store experience—especially for discovering new things. It’s one of those interesting tension points that challenges some of the common myths about different demographic groups.
Q. Considering the volatility in customer behavior, how do you approach predicting the future when you can’t rely on historical data? How do you handle questions that lack data for reporting?
A. Usually, my snarky answer to this kind of question is, ‘I don’t have a crystal ball.’ But honestly, you’d be amazed at how often I think back to my experiences during COVID. At that time, we constantly faced questions like, ‘What’s going to happen?’ or ‘Can you tell us what’s going on?’ It was a lot of pressure.
For us, the focus is on taking a measured approach. Instead of trying to predict one specific outcome, we look at a range of possibilities. Within that range, there are always going to be faster movers, slower movers, and those who wait to see how things play out. The key is to think broadly about what might happen and carefully consider the implications for our business under each scenario.
The danger comes when you confidently declare, ‘This is what’s going to happen,’ because the odds of being completely accurate are pretty slim. Instead, it’s better to present a spectrum of possibilities and focus on the ‘what ifs.’ What are the next steps or implications if any of these scenarios come to pass?
I think this applies to every business. It’s easy to get caught up in the immediate challenges right in front of you. But taking a moment to pause and think about the bigger picture—what might happen and how to prepare—is crucial.
For me, that’s one of the most challenging aspects of this role, but also one of the most rewarding. When the information you have is incomplete, you have to make a leap of faith. You base your decisions on the data you do have and your best understanding of things that can potentially happen.
