Me, Myself, and AI Episode 207

No Need for AI Unicorns: PepsiCo’s Colin Lenaghan

Play Episode
Listen on


Artificial Intelligence and Business Strategy

The Artificial Intelligence and Business Strategy initiative explores the growing use of artificial intelligence in the business landscape. The exploration looks specifically at how AI is affecting the development and execution of strategy in organizations.

In collaboration with

More in this series

Colin Lenaghan says he wakes up every Monday morning looking forward to the week ahead and what he’ll learn as he continues to lead digital transformation and artificial intelligence projects at work. With nearly a quarter-century under his belt working in revenue management at PepsiCo, these technology implementation projects keep him and his team on their toes while positioning the consumer packaged goods company for continued success long into the future.

In this episode of the Me, Myself, and AI podcast, we speak with Colin about some of the AI projects his team is working on and get his take on the skills and competencies organizations should foster to set up technology implementations for success.

Read more about our show and follow along with the series at

Subscribe to Me, Myself, and AI on Apple Podcasts, Spotify, or Google Podcasts.


Shervin Khodabandeh: Is there really such a thing as an AI unicorn? There might not be, but for sure there are horses for courses. Find out more when we talk with Colin Lenaghan, global senior vice president of net revenue management at PepsiCo.

Sam Ransbotham: Welcome to Me, Myself, and AI, a podcast on artificial intelligence in business. Each episode, we introduce you to someone innovating with AI. I’m Sam Ransbotham, professor of information systems at Boston College. I’m also the guest editor for the AI and Business Strategy Big Ideas program at MIT Sloan Management Review.

Shervin Khodabandeh: And I’m Shervin Khodabandeh, senior partner with BCG, and I colead BCG’s AI practice in North America. Together, MIT SMR and BCG have been researching AI for five years, interviewing hundreds of practitioners and surveying thousands of companies on what it takes to build and to deploy and scale AI capabilities and really transform the way organizations operate.

Sam Ransbotham: Today we’re talking with Colin Lenaghan. Colin is the global senior vice president, net revenue management, at PepsiCo. Colin, thanks for dialing in today from the U.K. Welcome.

Colin Lenaghan: It’s my pleasure, Sam. It’s an honor to be with you guys today. I’m excited to be in such esteemed company — really looking forward to the session.

Sam Ransbotham: We’ll definitely keep the “esteemed” comment in there for sure. Colin, can you tell us a little bit about your current role at PepsiCo?

Colin Lenaghan: My role, globally, very much is around building capabilities that help our business units win in the future, and there [are] three major legs that I try to push or that I am pushing. One is solidifying our foundation. That’s like, “Have we got the right talent in the right roles at the right seniority to integrate all the different elements that have to come together for revenue management?” The other leg, then, is, “Are we building for future growth?” So if you take the potential of our brands and how we can position them with the consumer as [they] relate to pricing and promotion, etc. — that’s a huge unlock for PepsiCo. And then this whole third leg is this digital transformation, we’re calling it. And that goes from everything from standardizing the analytics that we want all our businesses to be looking at across the world — to speed up diagnostics, to speed up decision-making and “solutioning” — right through to this more advanced AI agenda, and we call that our AI program. So it’s those major three areas. And, clearly, we sort of orient it with what the business need is across the world and what are the capabilities that we need to help and support to deploy, and clearly work to implement those and hopefully extract the value out of them.

Sam Ransbotham: What is the scope of this, that you’re working on?

Colin Lenaghan: The scope is right across the full spectrum of the PepsiCo portfolio — everything from beverages to snacks, of course … dairy. This is a classic capability that can help to solve some of the problems that we’re facing. So I think the categories. … We’re almost agnostic on the category. If there’s a really clear use case of where this capability can help us address a business problem, it really then can go across wherever we want. And I would see this applying as much to Quaker in the U.S. as our Quaker business in China, or our beverage business in Latin America to the U.K., right? It’s everywhere.

Sam Ransbotham: I think you’ve been at PepsiCo for 23 years, if I remember it right.

Colin Lenaghan: I’m 23-odd years at PepsiCo. And I actually started what you might call a net revenue management transformation in the U.K. eons ago, it seems like now. And at that time, we were building capabilities on systems, but very much analog, very much linear. I would say, literally in the last 18 months, what we’ve had to adapt to has been very rapid indeed. PepsiCo clearly sees advanced analytics, AI, as a core to how we’re going to have to operate in the future.

Sam Ransbotham: So, Colin, if you can, give us one specific example of a place you’ve used artificial intelligence. What’s one project you’re excited about?

Colin Lenaghan: I think the example is very much related to pricing — around how do we take very high level, very average pricing insights, and transform that. … If I could give the example, operating with 60 elasticities that help you understand where your pricing opportunity is to 40,000, right? I mean, that’s the sort of scale that you’re getting to. And then the decomposition of that elasticity around what it drags and draws from across your portfolio, across the portfolio in one retailer versus another retailer, really is sort of mind-blowing around what it can do. And that’s a real-life example of a product that we’re scaling up as we speak. Capabilities like that give us insight around what the right bets to make are and how we can journey toward an end state around what that portfolio price architecture could look like. I think if we didn’t have this type of capability, you would be flying quite blind. I think you’d be spending a lot of time with consumers trying to get answers out of them that they’re maybe not equipped to give. And I think you’d be taking bets and you’d be taking some risks because, as we know, pricing is the most powerful lever in a P&L, without doubt — stronger than volumes, stronger than COGS [cost of goods sold], whatever. You get it right, it’s amazing; you get it wrong, it’s devastating. Across all the things we’re trying to do, this capability on pricing, for me, is by far the most exciting. If we can get that deployed, let’s say to 15 markets in a short period of time, I think that becomes really powerful in our ability to understand the marketplace and to, candidly, make sure we’re giving the consumers what they want.

Sam Ransbotham: Yeah, makes sense. But going from something like 60 to 40,000 — that seems daunting. There must be a pretty big carrot out there for you to want to do that.

Colin Lenaghan: I think once you begin to see — and of course, the 40,000 means you take up pack to a certain customer, to a certain geography, right? It’s just a full deaveraging almost of your pricing. So that’s how you get that level of granularity. You know, my mind [has] now moved on to “This is now what is just going to happen everywhere,” because why wouldn’t we have this? If my partners in the data analytics team can tell me that we are industrializing this as we speak, you’re going to be able to spin this off and build it in weeks versus months. It becomes a bit of a no-brainer as to why we shouldn’t be considering that.

Shervin Khodabandeh: I wanted to ask, Colin, what’s the best part of your job?

Colin Lenaghan: The best part of my job? It’s almost a little bit why we’re doing this podcast. I honestly learn every day. That’s not a cliché or thing that I’m just saying. I mean I keep saying to people, “I still get the butterflies every Monday morning where I think, ‘Crikey, what’s going to happen this week, and what am I going to have to learn?’” And also, candidly, we’re building a capability with PepsiCo that’s relatively young, and I take great hope that there’s a legacy there that I can be proud of and that I’ve made a small contribution to helping the business. That’s what really excites me about this role and, as ever, with PepsiCo.

Sam Ransbotham: I’m going to give you a chance to redirect, though, from “Crikey, what do I have to learn?” to “Crikey, what do I get to learn?”

Colin Lenaghan: No, honestly, it’s absolutely true. It’s absolutely true, because these things are coming at me now. I have to embrace them. I have to learn. I have to be super uncomfortable. And honestly, I just go with it, saying, “Look, this is making me uncomfortable,” and that’s probably a good sign rather than a bad sign. I am going through a personal journey with AI, and it’s real and it’s eventful, shall we say.

Sam Ransbotham: So how are you doing that learning? I’m sure you’re a little further on the maturity curve than you’re admitting here. I think you’re being a bit modest, with your background.

Colin Lenaghan: I think there’s a couple of levels here. One, PepsiCo is very much an organization and a culture that learns by doing. I think we’re very targeted on key use cases where we see value for this type of capability. I think we’re operating very collaboratively with agility across how we get those use cases developed, how we prove them out, and then how we begin to scale them.

But then broader, I would say PepsiCo is making quite an investment in just bringing literacy of advanced analytics across the broader community, starting with the senior management. This is clearly something that has to be driven from the top. It needs cultural change. And so we are starting to do a lot of work in elevating that literacy — really getting the understanding of what this is. What do I need to know that is enough for me to embrace and leverage the capability versus be the expert, right? I mean, obviously we’re hiring a lot of amazing people to help us do that and partnering with a lot of great parties outside of PepsiCo to help us do that. I would say that’s a big initiative right now as we start to build more and more use cases and start to build more products, as we call them. Then, how we operate with that and, candidly, how we learn to live with it is a big opportunity. And I would say we’re all learning — I can speak for myself for sure.

Sam Ransbotham: Actually, the problem with all learning is that everybody has to learn. So you mentioned the senior management — how do you get, for example, those people to learn something that they, too, didn’t grow up knowing?

Colin Lenaghan: I think there’s a lot of factors that come to that. One, being very overt about it and taking time to make it important. The tone from the top, I think, is very important at PepsiCo. We’re all-in on this one. I think this is one where it’s kind of table stakes. There’s no debate. You’ve got to be really leaning in heavily here if you’re going to want to compete in the future CPG industry and broader. So I think it’s a combination of factors, and I don’t foresee it being an easy exercise or something that we’re just going to walk up and do a course and, hey, we’re all sort of literate. I feel this is going to be such a cultural process. Even how we’re working and doing things is much more fluid, much more a process and an evolution.

You know, you have to learn to really go in a very fluid way here, and there’s a lot of zigzag left and right. There’s a lot of two steps forward, one step back as you’re experimenting and trying to get this capability to do what you want it to do. So I have to say, personally, I feel quite uncomfortable at times. I feel I’m not in control. But then I’ve just got to look at and sort of engage with the team that we’re working with, and they kind of know what they’re doing, so you’ve got to have a lot of trust in this thing around that. It’s something that we’re building for the long term. And I think that long-term perspective is also very important. I think another point related to how are we going to learn — I think, seeing this not as a very short-term fix, that it solves an immediate problem in 2021. This is going to just be [a] capability that morphs into helping us solve lots of problems in the longer term. And so I think that’s why we view it as a very strategic capability helping us to solve strategic problems that, hopefully, over time, we improve, we enrich, we get better. And we strengthen the capability and, hopefully, the culture at the same time.

Shervin Khodabandeh: Colin, I want to build on a point you made about the process not being set in stone and the process itself is ever changing and sort of the new way of working between all the parties that are involved. What do you think the role of yourself as a leader and other senior management is in giving comfort and stability in a situation? As you said yourself, you yourself aren’t always comfortable with it. How do you guys deal with that kind of a — I would say — a culture shift?

Colin Lenaghan: Like many large organizations, proving stuff out and very overtly socializing that clearly builds confidence, builds trust, and helps you to move from one step to the next. Another aspect of what we’re trying to do with AI is the whole promotional optimization piece. If you take the results that we’re starting to see from these capabilities, it immediately inspires trust and credibility. And therefore that’s a really important process for us, because you’re starting small, you’re building confidence. And I think that’s going to be important, because if you really took on quite a large, big-bang approach with this thing, I just think it’s overwhelming, right? Like, I’m acclimatized or acclimated, as you guys in the U.S. say. I think we’re very much in to prove things out, embrace, of course, and be very aggressive in what we’re trying to do, but make sure we take the steps to take the organization with us to build trust and evolve the culture in that way.

Shervin Khodabandeh: This sort of process — that’s not necessarily always set in stone. Adaptability is part and parcel of getting these things to scale, right?

Colin Lenaghan: For sure. This capability needs to trickle down into the hands and minds of people who do the day-to-day operations; that’s where the real hard work is. So as we put task forces together to get, for example, an MVP up and running, that’s quite exciting, right? Because everyone’s there; it’s a project. But then you go, “OK, wow; that MVP worked really well. Now we’re scaling this thing. We’re taking it broad-based” — maybe to a category within a market or a customer, or whatever the case may be. You very rapidly get from a very excited project team to saying, “Hey, I’ve got to use this thing.” And then I think you’re into the classic potential areas where you run the risk of it becoming tactical versus it being incredibly strategic. Do people really understand their role [and] how they’re supposed to embrace it? I think for me in my agenda that I have within PepsiCo, those are areas that are very top of my mind right now. Like, “OK, how are we really going to operationalize this day to day?”

Sam Ransbotham: Well, let’s build on that. You mentioned more of the strategic nature of this. Some of these things seem like small changes. How do you keep this collection of small changes from ending up in a place that isn’t where your overall strategic focus should be? How do you balance that set of small changes and making incremental improvements with not just ending up at a place that’s slightly better, but wanting to end up at place that’s strategically better?

Colin Lenaghan: That’s a great question, because you can very much, in the day-to-day of these projects and what you’re trying to do, get lost in that, right? I think the use cases, though, that we’re trying to drive at — and in PepsiCo, in my space, those are very much around big levers for how we accelerate the top-line growth and improve the quality of our growth — pricing is super strategic, right? That’s not going to go tactical anytime soon. Promotional investment and using promotions to underpin category strategy and what we’re trying to do with the joint value creation with our retailers, that is super strategic and that’s not going to go away anytime soon. And ultimately, when you think about [it] from the user [perspective], is it always going to be efficient for these capabilities to be in verticals and in silos, or at some point, are they all going to come together so the revenue manager in the field has all these integrated at his or her fingertips that allows them to plan strategically, leveraging this capability? And that’s a little bit of the vision of where we might want to be going in terms of these capabilities.

This is very long term. I think it’s a little bit of the Irish saying that we have around here around horses for courses. And I think we’re quite pragmatic about that. But when you see the potential of it — and what I’m fully expecting is that more and more parties begin to embrace this — it then becomes almost an expectation of both sides that we’re going to have to operate with these types of capabilities. But if I take how we’re trying to set these things up, what we want them to achieve, could there be some longer-term vision of a more integrated solution underpinned with this connectivity and AI which it can do? That keeps certainly what I’m trying to drive super strategic, I think.

Shervin Khodabandeh: As you think about talent — I’m assuming those are probably the three most important dimensions: technical, organizational, and commercial — you’re finding that there [are] sort of unicorns out there that have enough of all three, or you’re finding that it’s required for everybody to have some of all three, but some would [have] spikes in some or others. Or could it work where you’ve got … talent that’s commercially and organizationally savvy on one side and then talent that’s technically savvy on the other, and combined, they sort of cover everything? I’m trying to get a sense of how much an individual contributor has to partake of those three main components.

Colin Lenaghan: It’s a great question. I think it also depends where you sit in the organization. If you take globally, candidly I need more distinction in the capability set. So I need a classic strategy-type individual to drive that strategy agenda, I need a capability expert, and then I need a digital/tech expert because we’re building capabilities. The more you go down the organization, the more you see the need for the blend to come together in the operating markets, and you don’t see too many unicorns. And I’m not sure you need to have the unicorn, right? Because ultimately, in the business, it’s around how you use this to commercialize and drive the business impact.

Sam Ransbotham: That’s good to hear, because I think it’s tough to find those people who can do everything. It sounds great on paper to find someone who does all those things, but the reality is … we call them unicorns for a reason.

Shervin Khodabandeh: It seems like Colin is saying you can build a unicorn by having the parts.

Sam Ransbotham: By having the parts?

Shervin Khodabandeh: Yeah. You’ve got the parts. Assemble them internally together and sort of create the organizational unicorn rather than [the] individual unicorn.

Colin Lenaghan: That’s right.

Shervin Khodabandeh: Is there a specific example of an area where you guys put a solution in place, and the organization went through that learning and that understanding, that you can talk about?

Colin Lenaghan: Yeah. I mean, we’re live in a particular market with one of those products that we have, and this capability is very much around how we predict a different shape of promotional calendar against the shared objectives between a retailer and ourselves. So it’s very much a strategic capability. For example, we might want to agree that, given the profile of the shoppers and the agenda of the retailer, they want to accelerate their top-line growth. In another retailer, it could be they need to manage the margins a bit more. And that’s taking a full category view of how this capability can help them to do that. And it’s not what you might refer to as common backward-looking tools. This is very much around forward-looking capability; against this objective we and the retailer have set, this is the shape of the calendar that needs to change for us to achieve that.

I remember we went through a great process through the MVP. … We probably took a month in virtual rooms or rooms or whatever, bringing all sorts of other third-party data, like panel data, like IRI data, to validate this crazy hypothesis that this algorithm was suggesting. There’s no way that [it] would work on this particular part of the category. And we managed to get sort of confidence that what the capability was suggesting for execution could work. And lo and behold, when we put it into the market — it was in the stores, hundreds and hundreds of stores — it worked. And this is a little bit about what we talked about before, around proving that use case out [and] showing how it can predict different scenarios that maybe we would be comfortable with. And that then builds confidence for us to keep this and use this as a more strategic capability.

Sam Ransbotham: Well, Colin, many thanks for your time today. You’ve, I think, brought out a lot of things that many managers are going to feel are urgent and are important, really for everyone — not just revenue management, not just snacks, not just sodas, but I think managers everywhere. Thanks for taking the time to talk with us today.

Colin Lenaghan: Thank you very much.

Sam Ransbotham: Shervin, Colin covered quite a few points. What struck you as particularly interesting?

Shervin Khodabandeh: I liked his comment about horses for courses, talking about how the new reality that [AI] creates in terms of how people will use it, how the big network of retailers and the whole ecosystem is adapting to [these] new ways of working, and how even internally within PepsiCo, teams are doing things in different ways — in more innovative, less static, more dynamic and test-and-learn ways. I liked how he talked about [that] this is going to be our future. We all know that, and it’s not going to be a one-time thing. It’s horses for courses, and that’s how the new course is going to be.

Sam Ransbotham: Definitely. Horses for courses makes sense because there isn’t just one horse that’s excellent at every racetrack. Similarly, I’m sure that companies would love to find an AI unicorn who is somehow magically good at everything AI-related, but it just seems unlikely. They’re not going to find [people] who are ML coding experts, who can talk fluently with business managers, and can communicate and present perfectly, and still [are] somehow willing to work at an entry-level salary. Instead, what Colin’s talking about, and like others are saying, success is a team composition, where different people bring different skills, and even if companies somehow could find a magical AI unicorn today, even that unicorn would have to learn, as technologies change so quickly.

Shervin Khodabandeh: Yeah. The other thing I really liked that he brought into the dialogue is, “Well, first of all, the best part of my job is waking up every morning and saying, ‘What am I going to learn today? And I’m looking forward to it, having been at PepsiCo for 24, 25 years.’” And I thought that’s really, really encouraging and energizing, coming from a leader.

Sam Ransbotham: It was a bit of a contrast, too; in some of our other episodes, like with Walmart and Prakhar Mehrotra, these people have gone through multiple jobs, and relatively quickly, moving from one technology space to another technology space, applying learning from different areas to their current role. Colin’s was a very different story. His own story about how he got to be in his role was quite different. He’s learned while he’s been there, not brought learning from other places.

Shervin Khodabandeh: Yeah. And I think that’s actually quite valuable in a place like PepsiCo, and it fits quite well to the point he made about what it takes to get these things to scale. It is about culture change, organizational change, and also commercial focus for value. And having been in an organization long enough to know what it takes to create that change, I think, really works for him. I could have imagined, if he came from outside with [a] series of very, very successful stints in highly digital companies, it might’ve been a very different trajectory for them.

Sam Ransbotham: So the other part there, too, is that with that ecosystem, Colin mentioned the butterflies that he got every day, but there’s a lot of butterflies in a lot of different people there, and getting that trust so people are comfortable with those butterflies [is a] hard job — and not a technical job.

Shervin Khodabandeh: No, it is a hard job.

Sam Ransbotham: Thanks for joining us. Next time, in our last episode of Season 2, we’ll talk with Elizabeth Renieris, founding director of the Notre Dame-IBM Technology Ethics Lab. Please join us.

Allison Ryder: Thanks for listening to Me, Myself, and AI. If you’re enjoying the show, take a minute to write us a review. If you send us a screenshot, we’ll send you a collection of MIT SMR’s best articles on artificial intelligence, free for a limited time. Send your review screenshot to


Artificial Intelligence and Business Strategy

The Artificial Intelligence and Business Strategy initiative explores the growing use of artificial intelligence in the business landscape. The exploration looks specifically at how AI is affecting the development and execution of strategy in organizations.

In collaboration with

More in this series

More Like This

Add a comment

You must to post a comment.

First time here? Sign up for a free account: Comment on articles and get access to many more articles.

Subscribe to Me, Myself, and AI

Me, Myself, and AI