In this episode, Luis Gonzalez, Global Supply Chain Director at Becle, the company behind the Jose Cuervo brand and the world's largest and oldest tequila producer, discusses how the company is using decision intelligence to transform its global supply chain. From dynamic forecasting to expanding into perfect order, Luis shares how Becle is building a more connected, data-driven value chain to support rapid global growth. He discusses the changing dynamics of the beverage alcohol industry, lessons learned from deploying decision intelligence at scale, and how Becle is improving forecasting, inventory management, customer service, and operational decision-making across more than 80 countries. Whether you're interested in AI, supply chain planning, digital transformation, or decision intelligence, this conversation offers practical insights from one of the industry's leading global brands.
Transcript
Fred Laluyaux (00:00)
Hello, and welcome to the Decision Intelligence Podcast, where we have real candid conversations with thought leaders, practitioners, and tech experts about AI, decision intelligence, and the future of work. I'm your host, Fred Laluyaux, co-founder and CEO of Aera Technology. Today we are joined by Luis Gonzalez, who's the global supply chain director at Becle, the world's largest and oldest tequila producer, and the company behind the iconic brands like Jose Cuervo. With more than thirty years of experience in supply chain transformation, Luis is helping shape how Becle uses decision intelligence to drive its business. Luis, welcome to the podcast.
Luis Gonzalez (00:42)
Thank you for having me, Fred. It's been great
Fred Laluyaux (00:43)
How are you?
Luis Gonzalez (00:44)
knowing you and working with you all of these years.
Fred Laluyaux (00:47)
And today we're finding you in Mexico, right?
Luis Gonzalez (00:50)
Finally, finally I'm here in
Fred Laluyaux (00:51)
Ha ha.
Luis Gonzalez (00:51)
my home actually. So happy to be here.
Fred Laluyaux (00:53)
Great. Well, before we start, can we talk a little bit about your background and your career journey? You've had an amazing career in supply chain, and you've seen multiple ways of technologies and transformation. So tell us a little bit about your career journey and what's changed most in supply chain along the way.
Luis Gonzalez (01:13)
Whoa.
Well, my career started more than 33 years ago. In the very production floor, I was working for Amoco Fabrics and Fibers in Mexico. I went as a quality analyst, then moved into production supervisor and later in process engineering. Then I moved into the apparel industry with VF Corporation, where I had my first international assignments, exposure to truly global
Fred Laluyaux (01:35)
Uh-huh.
Luis Gonzalez (01:36)
supply chains. Then I later joined Kraft Foods, and then I realized that I was now working for Mondelēz because Kraft had decided to split into Kraft and Mondelēz. And in that area, I take on leadership roles in planning customer service, logistics, and supply chain transformation before ultimately joining Becle, where today I lead global supply chain. And regarding your question, over the years, I've had the opportunity to work across five different countries in our industry. What has been particularly fascinating, and I believe you and I had a nice conversation this year about it, is the evolution of technology firsthand.
When I started, we were working with relatively basic AS/400 mainframe systems, right? Today we are using cloud-based platforms like S4HANA, Oracle, and AI-powered decision intelligence tools. So decisions, Fred, are certainly made faster now. The world has become infinitely more connected and complex. But there is something I have learned along the way, right? Faster does not necessarily mean better.
Do you agree? So in
Fred Laluyaux (02:36)
Yep, a hundred percent.
Luis Gonzalez (02:38)
that sense, we're still trying to solve the same fundamental supply chain problems that the company faced decades ago when I started. Only now in a far more volatile and unpredictable environment. As Gartner describes, we live in a VUCA world referring to increased volatility, uncertainty, complex and ambiguous work, but still we are dealing with the same supply, demand, and inventory balancing problems.
That's why in my career, if I reflect about it, I can tell you that this is an exciting moment for our profession. Technology in the past has helped us to digitalize transactions, generate more information. Now, with decision intelligence, we have the opportunity to move beyond simply producing data and focusing on understanding the consequences of our decisions, evaluating trade-offs, ultimately making better business choices. To me, that's where the real transformation begins.
Fred Laluyaux (03:24)
Talking about going from data-driven to decision-centric, right? This is pretty much what you're describing is how do I enable the decisions to be made better, more accurately, closer to the point of impact, as in real time or as close to impact, but also looking at how it impacts the entire value chain. So I like this framework of moving from data-driven, which means I'm giving people data and tools, to decision-centric, which means I start with the decision and I start pulling the string.
Before we go deeper into a conversation, tell us a little bit more about the world of Becle. People know the brand Jose Cuervo; most people do, but not necessarily Becle. So tell us a bit about the company and its global business, because there's a lot more than one brand behind Becle, right?
Luis Gonzalez (04:07)
Yes, absolutely. Becle, for everybody to know, is the parent company of Jose Cuervo. Jose Cuervo is the tequila company that has been producing tequila for more than 250 years.
And you and I had this conversation in the past – Jose Cuervo, if you think about it, has been producing tequila before even Mexico existed as a country.
And, you know, it's far more than Jose Cuervo. We have iconic brands like 1800, Reserva de la Familia, Tradicional, and now it's our tequila, which its demand is exploding in the US, which is Gran Coramino. But it's not only about tequila. We are producing whiskey.
We have the second and fourth largest brands of Irish whiskey with Bushmills and Proper Twelve, and we have strong brands of American/Canadian whiskey like Pendleton and Stranahan's. So it's a very, very diverse spirits business which operates in 92 countries with its own operations in three different countries because the nomination of origin: Mexico, Ireland, and the USA. And above that, we have some external manufacturers to better support the needs of our customers across Australia, UK, and other countries. My role here is leading the global supply chain, including end-to-end planning, logistics, transportation and warehousing, plus
Fred Laluyaux (05:27)
Yeah.
Luis Gonzalez (05:27)
trade.
And obviously customer service, which we have a great opportunity to better evolve our system and connect it all from an integral platform.
Fred Laluyaux (05:37)
Well, that's a very large, complex, but talk about a legacy business. Your brands were where you were making the product before the country was defined. So that's quite remarkable. So
Luis Gonzalez (05:49)
Yes.
Fred Laluyaux (05:49)
Yeah, legacy but very modern and very aggressive and very active. So you must have a very strong culture inside the organization, I'm sure. But let's talk a little bit about, you know, the market and how it is changing. So
Becle is changing; Becle is growing, going into other areas. But your market, the alcohol market, has changed dramatically over the last few years. So how do you describe that change and how did that drive the way you think about planning, forecasting, and working with your distributors around the world?
Luis Gonzalez (06:20)
Yeah, just let me give some context. Becle is a 2.5 billion
Fred Laluyaux (06:24)
Yeah.
Luis Gonzalez (06:25)
company. But in our case, what we are selling now has to be seeded in the fields more than 12 or 14 years ago. The agave cycle is a five-to eight-year cycle. Plus, in some cases, you have to wait for one, two, three years in the barrel. Same thing happens with whiskey. So in our cases, the complexity and volatility
of the markets have to be balanced with a long-term view of the planning challenge that we have.
Fred Laluyaux (06:52)
Yeah.
Luis Gonzalez (06:52)
Because, you know the demand can change. However, what I have in barrels, aging, cannot change anymore, you know, and for me, planning for our planning team, the planning cycle is obviously cycle by cycle, one month by month. However, the planning for us, the long-term planning for us goes for eight, 10, 15, 20, and 30 years. That's a different context.
Fred Laluyaux (07:17)
Wow.
Luis Gonzalez (07:18)
Right. Yeah.
And like most industries, we've seen complexity increase sharply since the pandemic. There is more geopolitical volatility, you know, tariffs shifting regulations across countries, lots of regulations with alcohol lately. And on top of that, consumer habits are changing. People are drinking differently depending on location. In other cases, we are seeing a quick evolution from selling off-premise to on-premise in the restaurants, and those patterns are evolving faster than we've seen in the past, and it's changing.
Fred Laluyaux (07:47)
What does that mean? Can you explain that?
Luis Gonzalez (07:49)
The on-premise versus off-premise. Yeah. On-premise, in bars, restaurants, now in developed economies are tending to drink less in on-premise in restaurants and bars and drinking a little bit more
Fred Laluyaux (08:02)
Got it.
Luis Gonzalez (08:03)
in their houses. In some developing economies, in some growing economies, it's happening exactly the opposite. But those changes that have been
Fred Laluyaux (08:11)
Yeah.
Luis Gonzalez (08:12)
very, very long cycles now are becoming faster review cycles, and in some cases,
Fred Laluyaux (08:17)
Yeah.
Luis Gonzalez (08:18)
we are not able to predict that with enough anticipation. Again, our production cycles can be very long for that purpose. So in that sense, we need to make sure that we capture the right data, the consumer habits, and transform that into manageable action plans and planning cycles, into our planning cycles. So that's where most of our ERPs, or the actual ERPs, are presenting a problem. If you think about the
Fred Laluyaux (08:45)
Yeah.
Luis Gonzalez (08:46)
configuration of ERPs, Fred, is that they are built on the one main assumption. You need complete validated data and a fixed number to operate it, right? The world does not act like that or behaves like that anymore.
So if the ERP needs complete data, fixed numbers, we need to start planning with more probabilistic approaches. In that sense, we have an internal joke. In which we say, we should design an ERP that can work with ranges because the demand will vary during the following years. So we need to adapt it quickly. And I believe that is an opportunity where decision intelligence can help us at the scale and the right speed because traditional ERPs rely on fixed models. And we need more speed and agility to react to those market condition changes.
Fred Laluyaux (09:38)
What's interesting also about your business, tell me if I'm wrong, but is the fact that your product as it ages gets more valuable, right? An older
Luis Gonzalez (09:46)
Mm-hmm.
Fred Laluyaux (09:48)
product is actually more valuable than a product that's coming off. So I mean, for some of your products. So that adds another layer of complexity, which is when do you release the product? What is that optimal point to supply, you know, the demand, but at the same time make sure that you optimize
for the revenue per unit of product. Is that a problem that you're facing?
Luis Gonzalez (10:09)
Yes, and it can be a happy or a sad problem because
Fred Laluyaux (10:12)
Yeah.
Luis Gonzalez (10:13)
the value of our products increases year over year when you have it in barrel. However,
Fred Laluyaux (10:18)
Yeah.
Luis Gonzalez (10:18)
the amount of working capital that that demands is humongous.
It's a real big amount of capital you have to have. So in some cases, several analysts tend to think all the time, and these are very high-margin businesses. Yes, but if you take into consideration all the investments and how many years you have to hold that inventory before selling it, you really have a different ROIC equation. So for us, the ROIC is a strategic KPI that we need to take a look at from the very beginning of the planning process. So that's what differentiates the spirits industry from other industries, I believe.
Fred Laluyaux (10:53)
Yeah, that's super interesting. It also gives you a competitive advantage because your business has been running for a long time. But you'd mentioned decision intelligence, so when you and I connected, I don't know how many years ago, but when you started on your journey, it was really a focus on the topic we're talking about, which is dynamic forecasting. And I think you've achieved pretty strong results. We've talked about this. So what did that first deployment of DI for dynamic forecasting teach you? And how did it shape your priorities for what's coming next?
Luis Gonzalez (11:23)
Yeah, well, believe it or not, forecasting at Becle historically has been an almost non-existent true global business discipline. There was little structure; whatever forecasting existed was fragmented across different markets, managed in silos, supported by different processes. For instance, Mexico was using
a small local tool while in Europe they were using something different. So there were different processes, tools, and even assumptions to run their forecast. So there was limited visibility into future consolidated demands. Planning decisions
Fred Laluyaux (11:55)
Yeah.
Luis Gonzalez (11:55)
were often reactive, and there was no common planning language across the organization.
So in that sense, after running several bidding processes, we were looking for a platform that could help us enable not only forecasting but other things. And we took a pragmatic approach, launched what we call Forecast as a Service model. You remember that, right? At the time, we did not have the scale and internal demand planning capabilities to build and sustain our world-class forecasting process across the global organization.
Fred Laluyaux (12:23)
Yeah.
Luis Gonzalez (12:23)
So we have partnered with an external team that operated the forecasting process while leveraging a platform as a technical
technology backbone. In parallel, we focus on building internal capabilities that now are taking over that planning process. So with that, we created a foundation for a more mature planning organization.
We started the pilot in the US, which represented the largest complex market. And you remember that I even called your team saying, guys, we are seeing really, really strong results. And we decided to start expanding before we had even completely stabilized the system because we started seeing
Fred Laluyaux (12:57)
Sure.
Luis Gonzalez (12:57)
results. And that was something that was really remarkable in the way we collaborated because we started
started expanding before stabilizing completely, but we eventually covered 92 countries. And after the first year, we were able to reduce the forecast error by half. So it was a really, really big success at the beginning. But it was not only about the way we produced the forecast and the forecasting error, but it really helped us to set the very foundation for the rollout of the integrated business planning that came after the successful rollout of the forecasting.
Fred Laluyaux (13:34)
Talk a little bit about the complexity. I know about it, but I don't think our audience does. The complexity of forecasting, just maybe in North America, right? With your model where each state you have to go through distributors. How does it work? How did you crack this nut? Because it's one that most people, I mean, as much as you can talk about it, of course, but this
is not a simple forecast. It's a very complex process that you have to go through with a lot of constraints, right?
Luis Gonzalez (13:58)
Yes.
Actually, what I can tell is as several other spirits companies, we are not allowed to sell directly to customers in the US due to the regulations, right? That's what we call the Bell States in which we there are some states in which we sell directly to distributors, and some states we have to sell through the Bellman States distribution centers, in which they distribute to the distributors.
issue here is that the demand signals and the
demand data is released in different buckets with different timeframes. And if you see the 52 states of the US and you start trying to create a unified cycle, it will be a humongous task for any human being. Because at the same time,
Fred Laluyaux (14:46)
Yeah.
Luis Gonzalez (14:47)
this industry changes a lot of versions. We can have seven, 10, 11 versions of the same products during that period of two to three years because we keep changing labels, or we keep adding some features into the same product. So when you try to forecast that using a human element alone, it's a humongous task, and it will always be human error-prone. What we had to deliver was a model that was flexible enough to capture even incomplete data and making the most after that data. And I believe that's
Fred Laluyaux (15:20)
Yeah.
Luis Gonzalez (15:21)
what the big advantage is. That you can work with incomplete data is better to have a portion of the data and to produce a forecast rather than waiting to have 100 % of the information and then you will be absolutely late.
Fred Laluyaux (15:34)
Yeah. No, it's a super complex model, because of different grains, different time horizons, data input, working with incomplete data, but you guys manage that beautifully. I'm so glad you got some good results.
Fred Laluyaux (15:45)
So, Luis, from those early successes with the forecasting, right? So you moved, and you expanded to the perfect order. So what does it mean, ‘perfect order’ at Becle, and how are you seeing decision intelligence skills helping your team deliver it? And what early results are you seeing with perfect order?
Luis Gonzalez (16:07)
Actually, the definition of perfect order is how to make the right product to the right customers at the right time with the right cost through intelligent automation with end-to-end visibility and fulfillment execution that is automated by definition. So we connect Aera with SAP to work collaboratively using our policies to make sure that we have that outcome in an automated way. In that sense, I was listening to one of the podcasts that you had some weeks ago, and were speaking about the high-frequency, low-value decisions, and you know this is high-frequency, low-value decisions, but as well high-frequency, high-value decisions as well. We are trying to automate both because that's where the value is in terms of servicing our customers at the end. For us, every single case has to be delivered to the customers, but we have to make sure they are sent right on time at the right cost as well across the world.
Fred Laluyaux (17:07)
Across the world. So you're doing this across the world.
Luis Gonzalez (17:10)
Yes, absolutely. Yeah, and that is one of the biggest challenges for us because, as you have to deal with multiple geographies, you have to have lots of manual interventions and unlimited end-to-end visibility. That's what we are trying to solve with Perfect Order because, as you pass from country to country and with different regulations, we have to deal with those restrictions in which we still have humans in the loop. However, my intention is to create the right policies, right decision frameworks for the system to work more and more autonomously. So we started first with the US, the biggest
Fred Laluyaux (17:52)
Yeah.
Luis Gonzalez (17:52)
region by volume. However, it's one of the least complex operations for us to test it. So big volume, lots of decisions that can be automated, but with a more, let's say, standard framework.
Fred Laluyaux (18:06)
Yeah.
Luis Gonzalez (18:06)
Now, after testing successfully in the US and piloting in the US, we are extending now to the jewel crown, which is Mexico. In Mexico, we tend to all the customers
Fred Laluyaux (18:17)
Yeah.
Luis Gonzalez (18:18)
directly from Mexico and going directly through Perfect Order. We are in that process right now.
Fred Laluyaux (18:24)
So, the reason US is less complex, is actually because you have this multi-tier model, right? Is it, or how do you explain
Luis Gonzalez (18:30)
This that's one piece.
Fred Laluyaux (18:31)
the US being more simple?
Luis Gonzalez (18:33)
Yeah, the US is more simple because it's high volume. Automated
Fred Laluyaux (18:36)
Yeah.
Luis Gonzalez (18:36)
decisions can be more standardized because you have 90% full container loads in some cases.
Fred Laluyaux (18:43)
okay.
Luis Gonzalez (18:44)
In Mexico, you have less than half of the container loads because you serve to small
Fred Laluyaux (18:48)
Okay.
Luis Gonzalez (18:48)
countries in Africa, in Europe, in several other countries. So that's a different operation. So the US has the big volume, low complexity situation in terms of the system. Obviously, we have all the complexities with the development states for sure in terms of the demand planning. But from the execution perspective, the US, given the high volume and the low frequency, is a different equation. It's relatively simple for a system to understand and to create the
Fred Laluyaux (19:15)
Got it.
Luis Gonzalez (19:16)
releases to serve it. That's why, right? And at the same time, you know, we are always looking for the biggest bang for the buck. At the end, we want to get the maximum benefit, and that's what the US provided. By connecting the available to promise to order lifecycle management in SAP to the TMS
Fred Laluyaux (19:31)
Thank you.
Luis Gonzalez (19:36)
transportation management system that we have with an external party, we are able now to not only have end-to-end visibility, but making sure that all the decisions that are being made are according to the policy that reduces the losses and the operational costs as well as maintaining the service level. It's a very interesting equation, but it's a complex one to create from scratch like we did, and really, really is delivering really nice results to the point that we are now expanding into the most complex markets like Mexico.
Fred Laluyaux (20:08)
And what results are you seeing? Can you talk about it again if you can?
Luis Gonzalez (20:12)
I would say that actually 60% of the decisions can be autonomously made, and in some cases we allow the system to run
Fred Laluyaux (20:19)
Wow.
Luis Gonzalez (20:21)
by itself based on the policy that we have created.
Fred Laluyaux (20:24)
Yeah.
Luis Gonzalez (20:25)
But as for the moment that we are in the industry, it is so critical in terms of making sure that we are warranting 100%
Fred Laluyaux (20:30)
Yeah, yeah.
Luis Gonzalez (20:33)
service to our customers. We are leaving
Fred Laluyaux (20:35)
Course.
Luis Gonzalez (20:36)
a little bit, so we still have some human-in-the-loop element in the actual tool as we will keep gaining more trust of the different sales teams.
We will be releasing more and more of the decisions without the human element, based only on the policies that we have agreed to have, you know. So that's a very interesting equation. That speaks about one of the biggest learnings in our journey, in which you have to closely collaborate with all the different pieces in the value chain, not only the supply chain, but
Fred Laluyaux (20:56)
I think that's quick and big; those absolutely.
Luis Gonzalez (21:07)
the distributor, the customer, in some cases, the development states, the sales team, the marketing team have to be involved in the creation of those policies before you set it up because once you set it up in Aera, it will
Fred Laluyaux (21:17)
Yeah.
Luis Gonzalez (21:19)
be automatically and disciplinary execution, you know? That's a very interesting one.
Fred Laluyaux (21:24)
I can't wait to update you on the new capabilities we're going to be releasing at AeraHUB around policy management, where the system literally understands and makes recommendations on how to improve your policies. And then you can take those policies and say things like, let me back-test a policy change on the decisions we made for the last two months in that specific area, or I want to run a potential test of a policy in the future, maybe in shadow mode for a while, to see how it would actually impact the way we're making decisions and ultimately impacting our our service levels and our customer satisfaction. So I need to update you on that because this is all new stuff we've been building, and we're releasing that very soon. So we'll need to make sure that we make some time for that discussion because I think you're gonna enjoy that. But I want to go back to a point you were making around the fact that this now is going beyond supply chain. It's going across the value chain.
At what point are you seeing, or did you see, this being more than a tech initiative and really starting to change the way Becle operates? The people, the process, not just the tech. When are you seeing that evolution?
Luis Gonzalez (22:31)
Actually, I believe I mentioned that in one of our conversations; there was an aha moment in the company in which part of the team came to us. They asked us for a specific meeting of the leadership team to present an initiative because after running the first cycles in EMEA and APAC, they were discovering that some items, very high-value items, like over $100, $200 per bottle. You know, we have some bottles that cost $4,000 per bottle. And so
Fred Laluyaux (23:03)
Yeah, I know.
Luis Gonzalez (23:04)
the demand was so erratic that it was not forecastable in the traditional point of view. So in that sense, by having the data available, the sales, the marketing team, and our demand and supply teams reached out to us, asked for a conversation, those guys, “based on the data that we're seeing after running hundreds of simulations, hundreds of data points analyzed.
We are telling you these items have to be treated separately. We need to forecast this separately because we cannot allow to lose a single case in terms of sales. We are not allowed. However, this cannot be forecasted in the traditional way. It should be not treated as part of the regular IVP process. So they came in with a solution already worked on with the finance group as well and said, guys, this is the value of the solution we are proposing, this is the working capital that we would need to implement it. But it was a really hard moment because only 18 months ago, we were discussing not having even data to produce the forecast. So that moment in
Fred Laluyaux (24:10)
Yeah. It's incredible.
Luis Gonzalez (24:13)
coming as a team and saying, guys, we need to change the way we operate. Made me think that it was no longer about us creating a demand planning process. We were enabling leaders to make business decisions. And I was one of the most valuable learnings across the process because when you implement a technology, you need to make sure you empower your team. And I believe that's one of the most powerful combinations. Empowering my team with the right technological solution
Fred Laluyaux (24:43)
Mm-hmm.
Luis Gonzalez (24:44)
for sure will deliver even better results.
Fred Laluyaux (24:47)
That's amazing. I'm such a big fan of what you guys have been doing and with a big sense of drive and, I would say, a sense of humility as well in the way we approach everything. It's been a real fun and pleasure to be working with you guys and with you in particular, Luis. First of all, congrats for the success that you've got in this so far. But now, you know, we have folks dialing in from all over, and when you think about supply chain in particular, you look at the future. What do you believe will separate the leaders from everyone else at this point, given what you know about the technology and its adoption? What do you think will separate the two categories here?
Luis Gonzalez (25:21)
I have maybe a different view from several other supply chain practitioners.
Several people think that the frontier is the way we enable the people, the way we run change management, the way we will run X or Y process, how we will encompass the new agentic AI with the actual ERP, things like that. But I believe the biggest challenge and the big differentiator for the companies that succeed at the scale
Fred Laluyaux (25:48)
Yeah.
Luis Gonzalez (25:50)
with implementing decision intelligence, and the rest is going to be that from the actual model, which you know, the models that we run, that we used to run our organizations as for monthly cycles, quarterly reporting cycles, things like that. But decision intelligence allows companies to run at a higher speed. And if you think about
Fred Laluyaux (26:04)
Yeah.
Luis Gonzalez (26:12)
it, most of companies are designed to run at a speed that the previous technology 100 years ago allowed us to. Now is the moment to rethink the way we run our companies to make sure we take full advantage of the decision intelligence and artificial intelligence tools that we have at hand. In other words, we need to redesign the whole company concept to make sure we take advantage of it. I believe that would be the big differentiator. Design the company now that we have higher speed in order to take advantage of that speed. I believe it is not a technological frontier it's an organizational design frontier.
Fred Laluyaux (26:52)
Right.
I love this. This is one of the reasons we built the company is to enable that change. And I would love to welcome you back whenever you want and have a deeper conversation on that specific topic. But we're running out of time. So I want to thank you so much, Luis, for the partnership and for sharing your insights today. Really, really lovely chatting with you. And I'm such a big fan. Anything you need, just, you know, ping me. You know, I'm not very far, and hopefully we'll see you at AeraHub.
I know maybe not, but your team will be there. So super excited. Luis, thank you so much.
Luis Gonzalez (27:24)
Thank you so much Fred. Really excited to work with you guys and looking forward to the new additions that you will release shortly. Thank you so much.
Fred Laluyaux (27:30)
I'll talk to you about it very soon. This concludes our Decision Intelligence podcast, From Dynamic Forecasting to Perfect Order: Becle's Decision Intelligence Journey. You can watch today's episode and find more decision intelligence resources at Aeratechnology.com. Thank you.