.png){kind=logo}
Fred Laluyaux:
It is now my distinct pleasure to introduce four senior executives who, like some of you attending this conference, are in the early stages of implementing cognitive automation. I'm personally eager to learn more about their respective vision, goals, objectives, potential concerns, and plans to overcome them. So with no further ado, and on behalf of all of us our Aera, please join me in welcoming Karen Jordan. Karen, you're senior vice president operations beverage division at Pepsi North America. Welcome. Will Beery, vice president of Global Transformation at Mars. Welcome Will. David Gutierrez, you're the CEO of DEACERO. I hope I didn't pronounce it too poorly. And DEACERO is the largest steel manufacturer in Mexico, right? With a global presence in Americas and in Europe. Welcome David.
And Roy Van Griensven, I tried really hard on this one Roy, corporate digital officer at Mitsubishi, Advanced Chemical. Welcome to all of you, such a privilege to have you guys with us today, and really exchange your thoughts with the community here. I know your time is precious, and I know you're very busy, so let's jump in. We know that the markets and the external forces have been challenging over the past several years, creating new threats, but also new opportunities. So I'm curious, how has this context changed the decision making environment at your respective companies? Karen, can we start with you?
Karen Jordan:
Sure. So first of all, thanks for inviting me to join. What I would say for us is it's really driven the need for, I would call it, speed and agility. So on the speed front, we've got to be able to make faster decisions. Things that we used to maybe meet about over sometimes months, mostly weeks, we now need to make those decisions in days, and then some cases, hours, just based off of, in particular, the environment that's been driven through this COVID crisis. It's really accelerated the need to make faster, better decisions. And another piece from an agility perspective, I don't know about other supply chain professionals, but we're in an environment in FMCG where there's not just one problem of the day to solve for. It's often many problems that tie into competing objectives, competing priorities, and in many cases, competing resources. So making higher quality, better informed decisions on a faster pace to really meet what I would call the ambiguous environment that we're in has become the highest priority we have in many situations to continue to move our business forward.
Fred Laluyaux:
So need for agility, the acceleration of the decision making, the pace of your decision making. Does that, does that resonate with you guys? Maybe Will, you're same industry. Do you have the same constraints and do you see the same evolution?
Will Beery:
Yeah. I think Karen said it well. I mean, Mars' culture is highly collaborative by nature. We're people oriented. So taking decisions traditionally is not the fastest thing we do because we like to align a lot of people around that decision before we move. And I think as Karen stated with the pandemic amongst other things, finding out who needs to make that decision and take in that decision faster is going to need to change how collaborative we always are. So definitely sensing the same thing that Karen described.
Fred Laluyaux:
Roy, David, you're coming from very different industries. Let's talk, well, from the SMCG. Maybe David, do you have any comments on the evolution of the context for you guys?
David Gutierrez:
Oh yeah. Thanks. Yeah. One of the things that we're working on. And I'm sorry about my English first, it's not the best.
Fred Laluyaux:
Don't worry. I can understand that.
David Gutierrez:
Perfect. Perfect. So one of the things that we're working on very hard is not only to align our people, which is something important. It's hard work that we're doing in there, but also to give them enough intelligent data for the decision making process so that they can be habilitated. I don't know if that's a good word in English or we can habilitate them for faster and better decision making. So I think that's one important thing that we're doing on our business.
Fred Laluyaux:
Very good. Roy?
Roy Van Griensven:
Yeah. I think what we see certainly in our industry is that we have tons of experience typically a lot of people in the company, not just in our company, but in the industry that know this industry. I think what the industry has started to realize is that the way that the industry has worked over the past decades is not per se the way that it will operate going forward as such, and the business models are changing, and et cetera. So I think the big change there that is really accelerating now is that the realization that you can't rely on past experience to make decisions for future, and it needs to be unbiased factual information and recommendations over kind of gut feel based upon 10 or 20 years of experience. I think that's something which has significantly increased, and that is in no shape or form a disrespect to all of the knowledge that we have. But yeah, the knowledge of the past is not per se what you need in order to make good decisions going forward because there is all the factors that play at this time.
Fred Laluyaux:
That's a very good point, and it kind of leads me to the next question, which is from you guys' vantage point, where is the leverage to improve the quality, the quantity, and the speed of the decision making to ultimately better serve your customers, which is what we're all trying to do. So David, love to hear your thoughts on this.
David Gutierrez:
All right. I'll try to answer this. It's a very good question. Let me just, before we start, to explain a little bit of our business, and I'll try to do it very, very fast. We are seven year old legacy steel manufacturing company. We produce and sell mainly commodity products. The company was founded in 1952, and since then we have always pushed ourselves in our investments also for a complete vertical integration. We began a process with sourcing and processing our own scrap, our own raw materials in our own scrapyards then we send that to our steel mills, and then to our wire plants. And finally we own our own distribution centers in to which our final customers. We sell and distribute more than 1000 families of products with more than 10,000 SKUs all throughout Mexico, Latin America, Caribbean, USA, Canada, and Europe.
David Gutierrez:
And one of the things that we do is we sell our products, all of our products, we sell them, deliver to our customers. So this makes for a very, very complex supply chain. So the leverage therefore, going back to the question, it comes from being able to fully connect our supply chain to achieve end to end, real time visibility that allows us for data driven decision. As I was saying, data driven decision making by using data analytics to sense and respond proactively to disruptions in the chain. Those allowing us to sense and react quickly to the markets, to the changes and reducing complexity in our value change management. This information, this connected information, not only allows us to better understand our customer and their behaviors, but also to be able to model our supply chain accurately to current demand, ultimately surpassing customer expectations. So just to give you an example, what we are targeting and what we are doing in Mexico is we want to supply all of our customers in less than 48 hours in all throughout Mexico, and in all the different market segments that we service.
Fred Laluyaux:
So in less than 48 hours. Wow. That's fascinating. Anyone want to add any color to the question? Anything you want to add? I think we touched on speed, we touched on agility. I like the point you made, Roy, around the fact that the past doesn't necessarily depicts to future, and you have to remove the biases. Those are the value that you can get. And David, you're talking about having this real time visibility and enabling your customers to be served in less than 48 hours. So when you think about this, I think we've positioned the need, we've positioned some of the value that's derived from this approach.
But what are the obstacles that you need to overcome to drive this effectiveness, to drive this agility in your decision making, right? You understand the why, we understand the how, but there are going to be some obstacles. And Will, you touch already on the fact that at Mars you guys are taking a lot of time to be very collaborative, and maybe that digitization needs to overcome that cultural challenge. We'll come back maybe to that. Roy, what is your perspective? And how would you... What advice would you have for the audience when they come to, "What about us? We have challenges."
Roy Van Griensven:
I think what significantly helps is of course the main obstacle, if any, is typically the human in this. And that's overcoming kind of a fear of being obsolete, not being recognized, losing kind of track of a kingdom, and et cetera. And now luckily, I'm going to say, in our company in Mitsubishi, we don't have that. So we've got proud people, but who very much understand that they need to drive on data to be able to make the best decisions in order to be reliable towards our customers, and et cetera. So luckily that human factor is less there, but certainly my past companies have seen that you need to just prove that for example, a statistical forecast is simply better than an army of planners can jointly produce altogether with all of the best knowledge that they have, and it's in no shape or form kind of a disrespect to any of the people.
So I think ones that human fact kind of is kind of gone, it's building upon proof, giving the people some feeling that, "Hey, the recommendations that are being made are actually valid and trusted." And then they start recognizing that, "Hey, we can better spend our time in more value added work and just let the system kind of continue. And you just need to rely on that to happen." I think the typical obstacle to focus on very early on, and that's something that we're approaching now is the quality of your data, of course, because in the end if you rely on data, the input of your data also to be very much correct. And don't forget about, let's say, also taking away obsolete processes because still you can arrive on data, but if you want to get that speed and agility, there are also many things that you simply can stop doing. You no longer need to perform that. Let the system run.
Fred Laluyaux:
Yeah.
Roy Van Griensven:
Accept the recommendation, execute it transactionally. And I think taking away obsolete processes, again, is a bit of a human factor, which typically comes a bit with cold water fear. But again there, find a pilot environment, build a trust that it works. And then the human factor is- [crosstalk 00:12:53].
Fred Laluyaux:
Yeah. I think you may raising a very good point. It's the start, stop, and continue. You have to really call out what you're going to stop as a result of implementing.
Roy Van Griensven:
Yeah.
Fred Laluyaux:
Will, you're further along in your journey at Mars who started a little longer ago. Did you have any of that, of these challenges and those obstacles that you had to overcome, and maybe you still are trying to overcome? Do you want to talk about that?
Will Beery:
Yeah. I mean, very similar story. I think the change management is by far the most complex problem in this journey. I think what Roy said about people being proud, it exists. We all think we do our jobs exceptionally well, right? So there's an element of, "Can you trust the technology to make a better decision?" But then there's also the humility of putting your pride maybe to the side, and execute it upon that. So it's a journey, and also holding that line. So once we do successfully make a change management call, it's not done and delivered. We have to constantly repeat and remind people, otherwise they kind of revert back to bad habits very quickly.
Fred Laluyaux:
Karen, I see you smiling a bit.
Karen Jordan:
Well, I agree with both what Roy and Will said. I would just say I was smiling because I also think there is something about data in data quality, data accuracy, data management, data integration that as you think about this world of cognitive automation becomes more important than ever, right? The number one question usually people have to get over is where did that data come from?
Fred Laluyaux:
Yeah.
Karen Jordan:
Do I trust it? And is it right? And so I do think there's a huge body of work. It's almost like to me, what's step zero before you get to the human factor, which is, it takes a very different level of commitment, structure, process, and management, I think, of data to really come out with value added solutions that people trust. In particular, if you're going from a world where the data is not naturally already on one ERP type platform or consistent, or you're trying to use data from varying sources that may not have ever spoken to each other inside of what David talked about earlier.
This concept of being an end to end supply chain, which is really kind of, that's the new buzzword, if you will. I forgot I've been using it for several years. So I don't feel like I'm behind the times, but I think this whole concept of data quality is almost like the foundation that many people I think have to, as you go into these systems, there's a lot of lessons there that can be equally as challenging as some of the lessons on the people front because ultimately people have been the ones looking at this data at times for years.
Fred Laluyaux:
Yeah. We talk about, at Aera, we talk about the four things we had to resolve when we thought about cognitive automation in the first place back when we started the company five years ago was data, science, process, and change. And you have to get the four pillars working in harmony. One of them fails, and the whole process of being able to let go, and delegate some of the decision making risk responsibility to a digital brain collapses. So we had to get the four things right. It took a quite a bit of an effort to get there. And in a previous panel, we discussed with two companies that have been... they are years into their journey toward cognitive automation, Unilever and Merck KGNA. And they were talking about the fact that the third thing that they do is exactly what you guys said, is fix the data problem.
Then you think about the... Once the data is trusted, then you think about the skills, the logic that sits on top of that data. And you start trusting that logic. And at that point, it gets going toward the self driving, and being able to take your hands off. But it's also fascinating for me from that vantage point that we talked to a lot of companies around the world. David, I remember being in a session with you and a large team of DEACERO executives. And you could feel in the room that the team was fearless. You guys were like, "We're going to go there, we're going to make it happen." What is it culturally that you make as a CEO that enables the team to be excited, and kind of looking at the glass half full as opposed to half empty even though the challenges with the data, the challenges with change are always there? What is it that you do to enable that culture of change?
David Gutierrez:
Thanks for saying this. Actually, I feel that we are still in the process of trying to make an ecosystem in our company that that drives innovation in motivation to do every day a better day. So thanks for saying that. I think one of the teams that you met, which is the team that is working on control towers is one of those teams that started with this idea of doing something completely new with... Actually the team was also brand new, and it's working very, very good. But one of the things that we're doing in the rest of the organization is, and I think some of you already talked... I think Will talked about this, but is to not only empower people, but align our people on what we are targeting as a company, and giving that responsibility to them to try to develop the result and the objective that we're working on. So I don't know, Fred, we're still on that process.
Fred Laluyaux:
Well, the fact that you always, it's a never ending process, I guess, right? You're never going to say, "We're there." But the fact that you are in the process is already amazing. So as we progress through the conversation, talk about some of the benefits of, and expected benefits of cognitive automat, right? Karen, you talked about the acceleration of the speed of decision making. You talk about the need for more agility. And obviously you are in a business where, if I'm not mistaking, people drink more sodas in the summer, and you have a very high seasonality, which in itself generates a lot of stress in your supply chain. So can you talk about how you're are planning to leverage cognitive automation to better respond to these shocks, to better absorb these shocks?
Karen Jordan:
Yeah, it's a great question. So we're pretty early in our journey from a planning perspective. One of the main reasons we're going down this path, and we're highly interested in it is because of the type of seasonality that you spoke about. And quite frankly, the seasonality is also magnified by the complexity of our portfolio.
Fred Laluyaux:
Yeah.
Karen Jordan:
So if you think about where consumers are today, consumers want more variety, they want more localization, they want more customization. And so the world where you might have a hundred SKUs to 200 SKUs has now magnified you 1500 to 2000 SKUs. And so some of our legacy processes around even using statistical models to create safety stocks based off of forecast variability, they're just not sensitive enough to be able to allow us to really meet the customer's needs in a consistent way, and do it in a way where we also can physically make, move, and store the amount of inventory we need to protect our service. So this is really about enabling us to be more precise, to be more accurate, and to be more agile about how we ultimately meet our customers. And ultimately, ultimately consumers desires for portfolio of many, many, many great products where the legacy models just... they're evolving, but we're looking to step change that evolution by using cognitive automation types of tools to help accelerate our business results.
Fred Laluyaux:
Yeah. Because your business was really built around mass production, mass distribution, mass communication. And it's changing toward consumers having strong appetite for customization and getting the products delivered to the door. So it fundamentally changes your, not just business model, but everything that you've been building as a company for the last decades, right?
Karen Jordan:
Yeah. So the way we talk about it is we want to be nationally great, and locally even better. So it's really this both in, right? You've got to be able to do large scale well, and we've also got to be able to do small well inside of many different supply chains, many different ways. We go to market through omnichannel solutions, and a very much more I would call it extensive portfolio that includes product formats that the tools that we built 20 years ago weren't necessarily designed to manage variety packs as an example.
Chilled products and cold chains, as an example where the complexity of the getting the supply chain requirements right every day means you just can't store 40 days worth of inventory to protect it. You've got to be much more agile or else you up throwing away a lot of things, and not necessarily meeting the customer's needs. So to your point, we've been working on these tools, we continue work on them, but we are really focused on how do we meet consumers' needs, and how do we ultimately support our customers, and having the right products in order to do a great job of-
Fred Laluyaux:
Awesome.
Karen Jordan:
... Of helping customers have our products.
Fred Laluyaux:
Roy, I saw you smiling when Karen was talking. Is what Karen's saying echoing such a different industry in your world?
Roy Van Griensven:
I think it is. And I think we'll probably all recognize many of the same thing, but I was also a bit smiling because sometimes we use the examples of... We have a quite steep grow target as a company, and I think a fantastic opportunity to grow. But sometimes we use the example, "Guys, we need to think about scaling as well, and think about if we ever become kind of as big as some of the FMCG, large companies, think about the Unilever, the PepsiCos and et cetera. You can't do what you're doing today while scaling to such a kind of volume. So you need to start doing other things." Then that's why I was smiling a bit. We use these examples to say, "Think about if we would be kind of that size."
And I think for us, it's mainly a question of how can we actually make sure that we enable the ability to scale where we're actually probably skipping a logical step because what we've seen is a lot of planning, a lot of analysis, a lot of insight based upon human experience, knowledge of the past, where the logical step was a little bit, "Okay. We need to prepare for onboarding a lot of data scientists where still people are doing the work, but then use technology to get insights. And then there is a huge amount of people that analyze the insights, and then come with good recommendations." And I think that's a step while we know we need to scale fast that that's a step we just need to skip. And I think this is an opportunity for us to skip that step to compliment, and to not step into, let's say, hiring a lot of data scientists, but actually start trusting the system to do that, the data science, and start relying on the recommendations that come out.
Fred Laluyaux:
This is why this panel is so fascinating. And somehow satisfying to me because we got two of the greatest FMCG companies in the world, but also chemical and industrial manufacturing. So it's fascinating to see that we're all here talking about the topic of cognitive automation, and how it can serve our different and respective industries. So Will, as I mentioned earlier, you started your journey a little earlier, and you embarked on a significant effort to digitize your supply chain with this multifaceted project one aspect of which you call the system of cognitive engagement. Can you talk a little bit more about what that is, and how you drive trust and engagement with that system? What's the cognitive engagement model that you've built together?
Will Beery:
Yeah, absolutely. Fred, your comment made me smile on scale drove demand, that's the from, and then the to being where consumer choice or behavior drives demand. And that's exactly how I talk about it both internally and externally for Mars. So when scale drove demand, we focused on manufacturing great quality products, highly efficient at scale because geographical reach is what drove your business for, but now that resiliency that was always there is being tested because the capability that we have today in mixed with the changes in demand are not allowing that to be as agile, if you will, as the resilience that it once had. So when I talk about it, I talk about changing from a sequential siloed supply chain from the factory to the customer, not even to the consumer or the pet parent in our case too, to one that's digitally interconnected.
So it acts like a complex system, and you manage performance across your supply chain as a complex system. So there's three pieces to that. There's making sure each piece has the most updated or best system of record it possibly can. That's one element of it. The second we've already talked about, which is then building the data and the analytics that make that visibility come to life. But the cognitive engagement for me is where you get the speed, the quality. Being able to see and actually act across the entire complex system as opposed to an isolation. Not really understanding what decision you're taking might be affecting in other aspects of the supply chain. So it's really that layer, that's the third, that we mean when we say cognitive engagement or systems of intelligent.
Fred Laluyaux:
Brilliant, brilliant, brilliant. So we talked about different things from the why, the how. Now you're getting into the nitty gritty of the models that you're putting in place. But if you think about it, cognitive automation is not just able to deliver the intelligence, the recommendations to your operators, right? So you get the data, you get the single data model. You've got the ability to deliver the recommendation and engage, but it's also capable of executing those recommendation autonomously, right? We talk about self-driving, we talk about hands off, and that's very specific about the technologies we're talking about. So with that in mind, when you think about the success of your current cognitive automation deployment, are you thinking about it in terms of internal efficiencies, or are you thinking about it in terms of the material impact on the core dimensions of your business or both?
Right. So there is a way to think of about implementing a new technology as just an internal efficiency, and we are going to measure it a certain way, or you're looking at a system that's more realistic that can actually do a lot of the work that you are doing today with your folks and your processes and your data. And if you think about how you measure the quality of the work, it's an outside in perspective, right? It's your customer satisfaction, it's your different financial metrics. So I don't know who wants to go first on that one?
Roy Van Griensven:
For me, for our company, it's very straightforward. It's always customer first. So that's where it starts. And I think everything, the speed, the quality all comes down to impact for our customers. Being a more reliable, faster, easy to work with company, to help our customers in the end, and to be the choice for them. Of course, it also provides a lot of internal operational excellence efficiencies, and et cetera. But for us, that is a secondary benefit almost that comes along. But it absolutely starts with customer first, customer benefit from battle decisions.
Fred Laluyaux:
Karen, Will any thoughts on that? How do you anticipate the measurement of the success of your initiative? What are the core metrics you're going to be looking at?
Will Beery:
Yeah-
Karen Jordan:
No, go ahead Will.
Will Beery:
Okay. Sorry. For Mars it's exactly actually what Roy described. So it was about winning with the customer and the consumer first, and doing that with high quality and fast pace. The efficiency internally at this point in time, we're addressing with some other capabilities, whether that's robotic process automation or other things, but I am absolutely interested in using the cognitive automation beyond just the KPIs or better decisions. I really think there's an opportunity there to truly transform and drive high efficiency beyond kind of the what's here and now.
Fred Laluyaux:
Yeah. Karen.
Karen Jordan:
I would triple ditto what both Roy and Will said. The only add I would say is for us internally, we're really looking at how do we reshape the work that our associates are doing in our facilities. So what I would characterize it as we hire really smart people. We bring them in. And many times some of the roles that we are looking at cognitive automation to help are jobs that we put people in the early stages of their careers. And so you imagine your best and your brightest come off of campus, and they're managing through seas of minutiae of data trying to figure things out, make decisions. And I think from my chair, it's the degree to which we can automate, I think, we will come out with better solution sets for our customers.
And I think for our talent internally, we will be able to better use the best of the brightest that we have to really add value in the places where we think we can get the most leverage from that human capital. So I think it is about the customer first, but the possibilities we see as how do we make sure we're using people's energy on the problems that most deserve their attention, not the problems that require their attention. And I think the automation offers new possibilities, and really being able to turn some things over to the cognitive automation, work benches and tools, if you will. And the freer associates up to work on the most complex problems that are probably the things they want to be spending more of their time, talent and energy on anyway.
Fred Laluyaux:
I'm so glad you brought this up Karen. We had a panel a few minutes ago with Professor Fuller from Harvard Business School who was also the co-chair of The Future of Work Initiative. And it's exactly the point he was making is there is a dimension, of course, of performance with customers and so on and so forth. But there is really a deep transformation on how work gets done, and the ability to attract the right and retain the right talent by focusing those talents on the core dimensions of where they can add value as opposed to what Joe Fuller calls dull work, which is the repetitive task, and that are just exhausting.
And I think this one year of COVID and made everybody realize that if you're going to spend all these hours doing work, you want to do work that has an impact, and that's rich, and leveraging your skills. So super, super interesting point. We're going to wrap up this panel. I'm so excited with this conversation. David, my last question will be for you, and it kind of echo echoes a little bit what we just talked about with Karen as well. As the CEO of a large organization, how do you think cognitive automation will change your company, and the way you work over the next few years? And I think it echos very well the point that Karen made a minute ago.
David Gutierrez:
Well, I think cognitive automation, it will enable us to become more agile, more flexible, more resilient by predicting customer and market changes, reactions and needs in a faster and precise way so that we can anticipate and act on their needs even before they know it. We're also targeting, and we will also believe that we will increase employee productivity by reducing time spent managing on day to day disruptions. We will empower and habilitate our employees for decision making. And ultimately many of these process can be automated completely. So we will spend less time looking for data, liberating valuable time to spend on tactical and strategic actions, and more time to perfect the way we do business.
We will also be able to have more synergies with our suppliers and customers. We will have an integrated visibility to align all the value chain that we will exceed our customer experience. And just to give you an example of what we're doing with the use of intelligent data in advance analytics, we have designed a complex assistance of raw material purchasing and production planning, and having able to optimize [inaudible 00:36:06] management, improve price setting to potentialize customer and supplier experience, optimizing the growth of the company. The implementation of the control tower, it will fully connect our data for enhanced cognitive automation.
Fred Laluyaux:
Thank you so much, David. I'm going to wrap up. I could spend another hour with you guys. Really enjoyed the conversation, and we're so excited to be able to get people connected, as I said, in different steps of the journey. Hopefully you enjoy the rest of the conference with us. Karen, Will, Roy, David, thank you so much for being our guest today, and for sharing your thoughts, and for taking a bit of time with us. Thank you. Thank you very much.
David Gutierrez:
Thank you very much.
Will Beery:
Pleasure.
David Gutierrez:
Thank you Karen, Roy, and Will. Thank you.
Karen Jordan:
Thank you. Bye.
David Gutierrez:
Thanks.
