Ferrero Rocher

How Ferrero USA Bridges the Planning-Execution Gap

Webinar Transcript

Introduction

Hello, everyone! Really appreciate everyone taking time out of their day to join us here. To kick things off, my name is Logan and I’m the Director of Client Solutions over here at Alloy. My team is really focused on working with our large enterprise customers to get them stood up on Alloy and make sure they’re getting value out of the tool.

We’re really excited to have with us Glenn Lawse today, who is the Vice President of Supply Chain over at Ferrero. I’ve been working with Glenn over the past 12 to 18 months, and it’s been a pleasure working with him and learning from him. Glenn is responsible for driving an optimized balance between service levels, distribution costs, and inventory through continually improving supply chain operations. He leads a team of supply and demand planning, warehousing & distribution, and customer service & logistics. Before Ferrero, Glenn spent more than 15 years at Johnson & Johnson in roles of increasing responsibility in areas of supply chain management, supply chain strategy, manufacturing, network strategy, ERP strategy, and supply chain & IT project and portfolio management. So again, really excited to have Glenn with us and looking forward to spending the next 45 minutes or so with you all, talking through how Ferrero was able to bridge the gap between planning and execution with Alloy.

In terms of our agenda, we’ll start with some more details on the problems Ferrero was facing when reality deviated from plans. Then we’ll walk through a specific example on how the process looks now, and then spend some time taking a step back and really reflecting on lessons learned and the ongoing journey to keep improving service levels, mitigate waste, and protect market share.

We’re definitely planning on leaving some time at the end to answer your questions. So feel free to ask any questions through the Q&A functionality in Zoom. We’ll make sure to leave time to address those questions at the end.

With that, I’ll pass the mic over to Glenn, and we can get things started.

Thanks, Logan. Thanks for the introduction. I’m happy to join you today. Tell you a little bit about Ferrero and talk a little bit about how we’re using the technology to help us.

Ferrero, your closest recognition may actually be the Ferrero Rocher pralines, which you see on the leftmost side of the screen here, commonly sold in many different configurations. But Ferrero actually also makes Tic Tac and Nutella, and recently we purchased a number of confectionery brands that used to be Nestle. So now, Raisinets, 100 Grand, Butterfinger, and Crunch are all part of our portfolio.

We’re headed towards about $1.5 billion in annual sales. It was $1.3 a couple of months ago when I looked at it, but the number keeps going up.

Got about 120 supply chain members on the team who are part of those functions that you heard Logan talk about – planning, warehousing, distribution, customer service. We have 10 warehouses we’re dealing with right now, and we have five different distributors. I say distributors because we sell both directly to retailers, but we also sell to distributors, who in some cases sell on to retailers, so we have a multi-echelon supply network down to our ultimate consumers.

Ferrero USA is a private company. It’s a family business founded in Alba, Italy, right after the end of World War II. It’s the third-largest confectionery company in the world. Massive footprint in Europe, for those of you who have had the opportunity to go there. We have distribution in about 170 countries with a big workforce of around a little bit more than 30,000 people across the globe.

Last year, Ferrero as a group closed its fiscal year at about 10.7 billion euros in sales. That’s more like 12 or 13 when you look at it in US dollars.

So we’re here to talk to you about the gap. Reality never matches the plan. If we were all in a room, I’d ask people to hold their hands up, but I’m sure this is something we’re all familiar with, right?

Certainly, the macro environment with the pandemic has had significant impacts, probably on all of our businesses, right? And we don’t really have much of an example or history to try to understand exactly what to do with that.

But the reality is, even outside of the pandemic, what we were looking to have Alloy help us solve is the fact that we don’t always get information fast enough or have visibility fast enough to make the right decisions. We have a lot of partners we work with. As you know, the supply chain sits in the center between all of your marketing, sales, and finance colleagues. They all have their own bits and pieces of information, and trying to make sure you make the best decision with all that information is sometimes very hard to achieve.

So, typically, you end up firefighting. You’re trying to pull inventory. You’ve got inventory, it’s in the wrong place. You’ve got inventory, but it’s the wrong code. The plans that you designed to sell one set of products turn out to have demand for another set of products.

All of those things are very typical, which I’m sure all of you encounter. Those are the kinds of things that we were really struggling with, and we were looking for a good solution.

If you think in kind of simplified buckets, between where we plan, what is it that’s on paper, for lack of a better term, in your ERP system, in your advanced planning system. We are looking at forecasted sales, forecasted shipments – forecast is all based on there being some sort of rationality, right? And sometimes it’s not so rational when you’re looking at what actually happens on the execution side, when you get actual POs and actual scheduled shipments. Obviously, you’re dealing with individual items on a customer level.

Sometimes in planning, depending on the horizon you’re looking at, you may be looking at aggregated demand and supply. We tend to look at planning at a national level, meaning we pay attention to certain customers, especially customers that might have a large portion of our business. But really, we’re looking at the national demand overall for our product, whereas execution really depends on what each individual customer is looking to do.

Ultimately, planning loves to hue, and we would all love to hue, to the rationality of consumption, but shipments don’t always match consumption for all sorts of reasons. Those reasons could be good, they could be bad, but there’s something of a reality.

Some of the common gaps when you think about this: We forecast a certain amount of sales, and yet you sell either over- or undersell that. There could be many reasons for that, which generally lead you to another set of investigations with data that you may or may not have direct access to.

Forecasted versus actuals is how it plays out. When you look at Weeks of Supply, Weeks of Supply mean something when you think about it in terms of your aggregated demand or your demand even on a SKU level, but it means something different to the various customers who may be carrying or have different levels of in-stock inventory or on-shelf inventory.

A lot of times, you end up in a position where, unfortunately, you don’t have enough to supply the entire market, so you have to make some decisions. It helps to be able to have some view into what your retailers have, what you have, and how you can make the best decisions to suit the overall marketplace, and ultimately to get the product in the consumers’ hands.

At the end of the day, consumption versus shipment is really the big bucket that you’re trying to solve.

This is a generalized version of how it looks from our perspective. Presumably, this is the same for many of you as well.

You have retailers or partners, some of whom may have sophisticated external tools – maybe pushing out their information on sophisticated tools, pushing it just to you in emails on a daily or weekly basis – and you have some retailers, of course, that aren’t that level of sophistication. The point is, it varies wildly, and they all speak their own languages.

The best you could do is hope that you have one or two people who are just fabulous at Excel and have nothing else to do all day but to stitch all this information together to try and make sense of it. Not only translating from one retailer’s language to yours – think about it in terms of Walmart has its own SKU numbers, Target has its own SKU numbers – you have to translate those into the internal ones so that your planners know what they’re looking at. But not only that, you have to do the same translation for multiple partners and try to understand across them; it’s a very difficult task.

Even if you had all of the Excel wizards you could hope for, it’s really a daunting task. And then Excel becomes that tool where you’re constantly looking to sell, to try and balance things. But the reality is all of your execution that you’re managing is inside SAP, so you’re disconnected from one piece to the other. To the extent that you do not have a tool to help cover this, it becomes inefficient. It’s very frustrating.

As you can imagine, these are the kinds of systems we’re dealing with. We have an ERP system that is largely fenced with our own information – our own manufacturing, our own DCs, our own forecasting. We have a consensus demand planning process. We have a data warehouse we use for forecasting. We do use some other tools around demand sensing to try and do the best we can internally. But most of those, frankly, are using historical information and the best cleansed information that we’re able to do inside our walls. It doesn’t represent something outside.

So what we’re missing is the distributor planning. A significant portion of our business goes through distributors like McLane. Obviously, McLane, as a distributor, has their own plans for inventory levels, for inventory management, and for stock management. They’re in turn looking at, talking to their customers who are your end retailers. That is not necessarily part of the same system.

And the retailers themselves have their own planning. So, for example, we actually do a lot of our business with Walmart via McLane. Walmart, of course, has its own planning systems as well.

So you have three separate planning systems before it leads to the point where you’re actually doing your production planning. It’s a perfect example of the bullwhip that can happen to you if you don’t have really good sync on that information.

Ultimately, then, you have a ton of reporting – you have BI reporting, you have Excel reporting, you have partner-specific reporting, you have POS reporting, but it’s all separate spreadsheets. At best, maybe someone stitches together two of them, once in a while, for deep analysis, but you can’t do it on a daily basis or a weekly basis. It’s simply not sustainable.

This should be something I think is common for everyone.

We like to have interactive conversations here, so as we go through, we wanted to simulate being in person as best we can. We wanted to do a polling question, so that should pop up on everyone’s screen now. We thought it would be interesting for people to plug in here. So if you take a minute or so and answer the question, that’d be excellent.

So not a surprising result. Glenn, you can see the response here.

Yeah. I’d love to know who that person is who only has one system; that’s awesome. And also, perhaps if you have any job openings. But as you can see, the challenge is fairly heavy here.

When we think of our end-to-end supply chain, the question really is, what is your end? What are you thinking is the end? Is the end when it leaves your facilities or your distribution centers? When you add in all your customers, that becomes a much, much bigger number.

I think that was illustrative of the kind of thing that we were challenged with and trying to deal with regularly.

So the glue is what we were missing, and we turned to Alloy to try to help us with this. We have been working with them for a little bit over a year now, on a number of our major customers. We don’t have full coverage, but we have a great starting point with about 50%+ of where our shipping points are.

It’s key that Alloy really allows us to bridge that gap between what we thought we’d do within our planning system and what was actually happening, but then get information to explain it. What’s actually happening in real-time in terms of key account forecast, visibility, and in-stock levels, down to the store level, so that we can see and be able to help not only have supply chain conversations, but sales conversations. It’s not purely a matter of saving inventory. There’s also a top-line conversation you can drive with this. Having that single source of truth, one place to go to be able to see all of this information, was really key for us.

We want to talk a little bit about what Alloy does for us and give you a sense of what we’re doing. We’ll give you a couple of generic cases that absolutely have a ton more detail in the background. In a way, we’re sort of making it sound really easy, what Alloy is doing here. They’re doing a ton of stuff to harmonize and ingest data in the background, and what Logan will talk about a little bit later.

At the end of the day, the purpose for me is to have better information to make a better decision. Or even have a better discussion. Sometimes I may not be the sole decision-maker on something, but it’s better if we can all talk about the decision to be made with my cross-functional colleagues having the same data in front of us, as opposed to just fairly typical conversations you think of in the past, where your marketing colleagues might have consumption and your sales colleagues might have in-store levels at your retailers, whereas your supply chain folks are looking at maybe some historical or statistical base demand forecasts. We want to try to bridge that all together.

Let’s take a look at what a couple of these examples are.

We had a new product launch recently, about a year ago at this point. A number of our products are actually not manufactured locally. They are, in some cases, manufactured in Italy, which obviously lengthens our supply chain and decreases our reaction speed significantly.

It has to come from Italy; there’s an ocean somewhere in the middle. Then it goes through a Ferrero DC. Then, in some cases, it goes to a distributor DC. From there, it’ll go on to Walmart – we’re gonna use Walmart in this specific example. From a Walmart DC, it may then go to a Walmart retail store.

As you can imagine, new products don’t have a history. You’re maybe modeling it after something else, and you have sales coming in to say “that thing’s going gangbusters.” We need to make more products. We need to push more product right now, and push more product with something coming from Italy, which could mean an airplane. So this could be really expensive at some point in time.

You really need to understand how I know whether it’s really a good decision to do it or not? Again, it’s not that you’re distrusting your cross-functional colleagues, but you really want to make it more on a view of data than a personal view based on a single person’s say-so.

Does this seem familiar? How often do you see this in your businesses? We know that demand planning is always wrong, but how often do you come up with a major question, especially one where you have to make a decision that’s gonna commit company resources, more than just, I just need a little more this week? It happens fairly frequently for us.

Let’s see what everybody else says.

Okay. Well, everyone’s between daily and monthly. Monthly seems like a nice place to live. It’s not the neighborhood I live in, but the reality is, everybody’s dealing with this constantly. Because most of the time, even when you have a monthly one, I’m willing to guess that the decision that gets made there is not a one-made decision. It’s multiple meetings, data, and investigations. To make the best decision, you really have to see what kind of data you can deal with – what kind of data you might have out there.

Let’s take a look at what we used Alloy to do in this example.

The graph you’re looking at is the capability that Alloy provided us, showing us actual shipments, which are the green bars. These are in units, so they’re sort of quantity agnostic in that sense – actual consumable cases. And the dotted line you see is actually Walmart’s forecasts.

So at the time, on March 2nd, March 9th, and March 16th, we were rolling into Easter, as you can see. We were getting signals that we had to make a massive increase in production, and that was going to drive a tremendous amount of extended cost. To a certain point, your safety stock can maybe suck some of that up, but at a certain point, you obviously have to increase production.

The reality was that when you even looked at Walmart’s forecast, it was dropping back down to a normal level after the fact. So this is a snapshot you’re looking at, that was taken right around March – somewhere between March 16th and March 23rd. You can see that Walmart’s forecast has a peak, and it does seem to be well-correlated with where their actuals have been in the past, but that is actually going to slide as we come down out of this period.

What actually happened? Here, the line is a little bit different. The line is not actually forecast this time; it’s actually how much of their pipeline they were sitting on – how much inventory the retailer was sitting on.

As you can see, we had to very quickly make sure that we provided them with more inventory after that big spike. Then ultimately, the numbers actually smooth themselves out.

Sometimes a single piece of information, which is something’s selling really well or, conversely, maybe unfortunately, something appears to not be selling really well, doesn’t really tell you the full story. You really have to bring several pieces of information together to be able to try and get the kinds of information to make a good decision.

Alloy gives us the ability to not only do that graphically, which certainly allows everyone to participate a little bit more easily in the discussion, but also put it in whatever units make sense to people. We can talk cases, we can talk units, we can talk – we have a weird unit of measure inside Ferrero that’s a weight unit of measure, something similar to kilos or tons.

We can also look at it by individual retailer, and you can see where the stories are of who has the single greatest amount of inventory. When you take a look at this in the table below, it gives you a little sense that Walmart’s pipeline is really quite powerful, but actually, they aren’t in too bad of shape right now in terms of what they’re sitting on, in terms of weeks of inventory.

Maybe we’re a little concerned about the amount of inventory that McLane has there. 35 weeks would seem to be like a very, very large amount of supply, especially when you’re dealing with a food item that doesn’t necessarily keep and is not designed to keep that long.

The ability to see multiple pieces like this is exactly what we’re looking to be able to do on a dynamic dashboard. And the thing that’s important to remember about this is: this updates generally as fast as your retailers can give you information. So in the case of McLane, Target, or Walmart, we can get it daily. We can see this daily without having to re-update the spreadsheet, re-do the downloads, etc. Daily usage to be able to track something, customize what you may be looking at, and drive a better conversation, either cross-functionally, internally, or potentially with your retail partners.

Some of the other capabilities we can do. You can share dashboards. Dashboards can be designed by a single person and be published to the rest of the group, or a select set of the group. You can share an individual version of a dashboard, much like bookmarks functionality, to say, “Hey, look, I’m seeing something over here, and I’m thinking I’m seeing this. Can you take a look? Are you seeing the same thing? Am I missing something?”

This is the kind of thing we want to be able to send to people. You could just pop it in their email; they see it the same way. If they have access to Alloy, they can link back right in and see the same charts that you’re looking at. They can drill down, drill up. They can change the time series or perhaps the set of buckets that they’re looking at things in.

The idea is that the context is easily there. It’s updated daily. I can customize it. I can drill down to where I think the problem might be and then quickly share that with someone I’m working with on the tool.

Another capability and functionality, which we really like and honestly, we have not explored enough – I’m teeing that up for Logan, so now he knows that the next time he’ll talk to me, we’re not exploring this enough right now, is the ability to set alerts for yourself.

As you can imagine, we’re looking at things that are overselling or underselling the forecast. We can set thresholds on that. We could set a generic threshold for the entire business, or we can set a threshold at an individual brand level, individual SKU level, individual retailer level, whatever seems to make sense for us. That will allow us to preset the things that we look at or need to pay attention to on the way in.

I can imagine all of you do some sort of planning by exception and probably wish that you could plan even more by exception than you do now, today. This is the kind of ability that helps us to be able to do that. Those alerts can be something that’s the first thing you see when you come into Alloy, you can be notified via email, or you can notify a group of people.

The functionality is there to help you make sure it brings your attention to what’s really important or what you’ve designed to be important. And if that changes the next month, you change what the alert looks like.

At the end of the day, you say, “OK, Glenn, this is really great. I like that. Those are cool. But what’s it doing for you? How is it helping you improve your performance, decrease your costs, improve your efficiency?”

Ultimately, you want to talk a little bit about what we’re doing to do this. And we continue to explore. It is a journey, truly. We’re not there yet. We have certainly achieved some benefit, but we continue to see more benefit and more use cases, and every time we see that, that increases our footprint and how we can really deliver on our efficiencies. Let’s talk a little bit about what that looks like.

Take a look. Here are a couple of key top priorities. You could pick a couple of items on this one. What are your top priorities?

I picked three only because I didn’t want to pick all eight. Let’s see what the results tell us here in a moment.

Aha, increasing forecast accuracy is the leader on this one. I’ll tell you, that’s the single, probably between that and reducing waste and spoilage – again, perhaps more of an issue for food companies that do not have a long lifetime in their products – but for us, increasing forecast accuracy and reducing waste are the key things we are going after right now. The ones we’re looking at for the next year truly are around reducing lost sales and improving on-shelf availability.

I think that we talk about some of this, but this is really a lot of supply chain terms more than anything else.

But there’s really one thing I would encourage you to try to think about as well: there’s an entire set of sales applications here that maybe it’s just not natural to our language in the supply chain. Because it’s not purely about a matter of, you see a lot of “reducing… reducing… reducing” here. It’s also about improving and adding to the top line as well.

That looks fairly familiar to me.

What do we want to do in terms of improving service levels? Improving service levels, certainly for us, is an important thing. I’m sure it’s an important thing for everyone.

The ability to understand inventory levels at our key customers, at the customers that drive more than 50, 60, 70 percent of our demand for any given item – the ability to have that immediate visibility into their inventory is one level beyond anything we’ve had in the past. Maybe in the past we’ve used IRI, maybe we used something else that told us POS generically at the end state. But to not be able to tell what’s happening on that entire supply chain, up and down that supply chain, really puts you at a disadvantage.

Having that visibility is super important to us. In fact, early on, when we were considering our business case for this one, my finance partner, who is naturally challenging me on the investment, said purely having this visibility alone, and having that visibility beyond just the sales team, is something that’s already helped pay itself off.

It’s not just inventory position; it’s also understanding your allocation. We are, for better or worse – if you are contributing to this, I thank you – Nutella, during the pandemic, has been selling like hotcakes. We have been selling more than we can make.

As a result, we are in a good position to make a supply-constrained decision about who’s going to get it. When we can provide more information, it really helps, so we can see even a better view and dig down deep to see: they’re all ordering, but who is really sitting on the supply and who has zero or major in-stock problems at the retail level.

Better manage product launches as well. We talked about this in the example earlier. The reality is, you’ve got products that you’re basing on some historical curve and some other product. We see that the information flow that comes back to you is not nearly fast enough in many cases to be able to make the right decisions to ensure that after the pipeline, as your POS starts taking off, you’re in the right position with neither too much nor too little.

What has Alloy done for that? Again, it’s helping us connect these workflows, helping us connect these inventory levels, seeing those real-time inventory positions. And it’s truly real-time, I have to underscore that. We may have had access to a single partner intranet daily to see something, but no one had the bandwidth to go in and do that work manually.

To be able to connect a little bit of the demand signal and store-level inventory position – and honestly, too, we tend to say Walmart like Walmart’s one place. Walmart has many, many stores and distribution centers, so the reality is:

We can sometimes say, “Yes, Walmart needs product, but only in these certain areas of Walmart.” Areas of the country, perhaps, because they’re the ones that have both the highest ring – the highest sales turn – and also the ones with the biggest inventory or in-stock problem. Setting those and using the alert capability and the email sharing capability have all been things that we’re using Alloy to help us with.

It’s not just about service levels, though, for us. As I mentioned, it’s about mitigating waste. We’re a food product. Ferrero Rocher, as you remember, the small pralines that are shaped like a golf ball, do not improve with time. They are not wine. The reality is, we need to move through those kinds of things very fast because we want the customer to have the best experience with it. Unfortunately, in many cases, they’re also made quite a long way away.

There are times when it is difficult to triangulate between what consumption is telling you, what sales are telling you – sales as in your sales team, who’s tasked with continuing to drive the top line, and what your actual shipments are telling you. It’s very rare that all of those line up perfectly.

The reality is, if you think about food, or many of you probably have seasonal products as well – we have some seasonal products, we make things the shape of an Easter bunny for Easter, we make things in the shape of a Christmas tree for the holidays – those items do not have a ton of demand the day after the holiday is over. It’s ideal to be able to make sure you’re monitoring that in and out.

Alloy is helping us there. Again, watching our POS visibility. Watching not only after the event has happened, but perhaps on the way into the event, saying, Where is it? Do I maybe have to talk to the retailer about taking some action to improve movement on the product beforehand? Can I offer a markdown or something in the last week or two because I don’t see that it’s moving well enough?

For us, this has been really important us to be able to achieve this.

It’s been critical to have the Alloy information to help connect those dots. Not only provide real-time inventory and turn information, but it also connects the dots about where I’m likely to have spoils. And can I do something? Can I work with that retailer earlier to move the product faster?

Maybe with sacrificing some amount of profit, of course, but at the same time, saving from having to destroy something at the end of it. Very important for us.

Obviously, these things don’t all start out on day one. Logan here is very capable. I did expect much of him on day one, but honestly, there’s just as much learning, and equally important, like any technology implementation, to align your people and process internally. That is a challenge that we’ve been facing, and we continue to work and drive on.

Change management is key. I don’t need to tell you this because that line could have been stolen out of any other presentation that you saw at Gartner or somewhere else. Change management is key.

Starting small and gaining traction, then expanding. This one is important to me because I think those of you who have been looking at any technology investments today, we’re in a good world for that. Because 10 to 15 years ago, there was no such thing as a small investment to try something, improve it, and then expand it. Think ERP. You couldn’t just do a little ERP. You had to do all of it, and there was a three-year payback. The idea here was for us to have very clearly defined use cases, apply the solution to those use cases, learn as we’re going, but also then continue to look for new use cases as we expand.

The other one was, honestly, bringing sales leadership in early.

I think this platform has more value to my company at the moment for sales than it does in the supply chain. It has tremendous value for the supply chain, but the supply chain value that I have tends to be more around “prevent,” and “prevent” is not a good business problem. You want to add or increment, which is truly a sales problem.

So those are the things that we have to do, and I think bringing them in early is a big deal.

I will tell you that this goes hand in hand with change management as well. You’re probably familiar with the fact that a lot of times, visibility makes people nervous. If you’re providing visibility to everyone for data that is usually controlled by only a small group of people, they feel a little bit nervous about the lack of control. They feel a little bit nervous about whether your data’s right. All of that you have to approach early on and find the key partners who can help you prove that case and prove that benefit because the benefit’s absolutely there to be had.

So I gave you a sense for how Ferrero is using this. I’ll let Logan talk a little bit more generically about the capabilities of the Alloy platform.

Thanks a bunch, Glenn. Again, really appreciate it. Now, before we get into Q&A, we did want to just spend a couple of minutes getting everyone a high-level overview of Alloy, so you can all get a sense in a bit more detail of how it works. Just one slide here for the group.

This schematic really summarizes how we think about the product. If we look here at the base, Alloy’s data platform collects data in real time, to Glenn’s point earlier, across lots of different sorts of data sources. Then we harmonize all of that data into a single model. We can get data from brick-and-mortar retailers. We can get data from eCommerce channels, direct to consumer, distributors, 3PLs, and this is all based on existing integrations and data relationships that Alloy has. We’ve got over 600 reusable integrations, and that number keeps growing. That’s in large part because it’s been a focal point of our organization since we’re really purpose-built for consumer goods brands.

Once all that data is captured from your ERP, from your point-of-sale data, from your distributors, and integrated, our modeling layer uses the data to create a network map of your supply chain. We’re modeling the relationship between your partners and their locations and your locations, ultimately, so that you can trace your products all the way from your production facility through your internal distribution centers, through distributor DCs, all the way down to stores at the point-of-sale. As I’m sure most of you can appreciate, modeling that complicated supply chain network isn’t an easy problem to solve. First, you’ve got to be able to model at the product level – lots of different product identifiers, potentially different case conversions, you have different product attributes and product categorizations, and that all needs to be brought together in a single data model.

Then you’ve got to be able to model that location network. This would include your ship lanes. It could be that, depending on the product, it goes through a different lane in your network. What do your lead times look like? Is it coming from Italy? Is it coming from far away or close? What are your location attributes to make sure you’ve got a meaningful network map?

Lastly, we recognize that time granularities matter a lot. Again, depending on where data comes from, you’re gonna have different fiscal calendars, you’re gonna have different fiscal weeks, different time granularities – ideally, we can get daily, but sometimes it’s weekly sorts of data. How do we make sure that we can compare apples to apples in that network over time? That’s really how we think about modeling this data.

That last layer in our data platform enriches that data once it’s all integrated and modeled. We’re able to enrich the data. We can compute and simulate metrics, give you out-of-the-box, dynamically calculated Weeks of Supply. We could show you where there’s potential for phantom inventory. At any point in your supply chain, we can project inventory shortages. We can illustrate to you the sales that you’ve lost from the product not being on the shelf, so you can better understand what the true, unconstrained demand is at the point of sale. Really, lots more of these sorts of insights that can help supply chain folks find and resolve the most impactful and important opportunities.

Plus, once all that data is together, we make sure we preserve it in a format that’s useful for any user. You can toggle from the Walmart language to your own nomenclature to McLane really easily, so depending on who’s consuming the data, we can make sure we’re speaking the right language, so everyone really can understand and take action as necessary.

Lastly, on top of that data platform, we’ve got two key applications for all this updated, integrated, and model data. We’ve got analysis, and this provides users the ability to customize dashboards, alerts that we’ve shown, and run experiments to deliver business insights.

Then we also have forecasting, which generates a demand forecast at the point-of-sale and translates it into a shipment plan. Hopefully, these two minutes we’ve taken here give people a little more meat on the bones to contextualize what Alloy is all about.

Although we do have specific solutions for sales teams and supply chain, we recognize the different use cases people will use the data differently, we are a company that really believes that you need a data platform, a single source of truth, to help bring together cross-functional groups to be able to solve problems because most issues and opportunities do require teams to collaborate effectively. I know we’re more than happy to dig into this more with folks on the call, but I do wanna make sure we leave some time for the Q&A.

We got one question right off the bat from Rob joining us. The question was, how did you get your IT team to work with you in exploring this technology?

I actually think that’s a great question because finance isn’t your only stakeholder in getting the money, right? In some cases, it’s around IT. I will tell you that I have a particular strategy that I follow in this. I worked with our IT team on this, and there were several common evaluations we had to do. Some vendor evaluations, some sort of cybersecurity evaluations, of course, but we went in saying no integrations to ERP. The minute you say no integrations with ERP, for most IT folks, they watch 95% of their work just disappear.

I do that very specifically for a couple of reasons. Number one, the fewer integrations I can drive and still start to achieve the benefit, to achieve what I’m looking for, the lower my opportunity costs on the way in. If I don’t have to spend $100,000 on some interfaces with ERP, and don’t let that number scare you, that’s me thinking of consultants. I know that Alloy makes these connections much simpler. If I don’t have to spend that, if I don’t have to engage them on that, at this point, they become more of a disinterested observer as long as the technology meets your information technology and security requirements, more than anything else.

I don’t know if that directly addresses your question. What I have done, however, is continue to make them part of the process where we’re evaluating or reevaluating or renewing contracts or adding capability. We are at the point this year where we actually may start considering some more direct feeds. We do provide a little bit of information – things like material master information – we provide ad hoc to the Alloy team on a monthly or as-needed basis because, honestly, while it does change quite frequently, in most cases, we’re able to survive with that being done manually. It makes more sense for it to be automatic, and between that and some of our inventory levels, things we’re going to explore more recently.

The simple answer is to start small because, honestly, if it doesn’t work for you, you haven’t spent nearly as much. Part of this is making a good business case and decision to try a technology, see that it works for you, and then evaluate as you continue to see benefits.

What I can tell you is this year, we’re just completing our renewal process right now, and we did a very robust exercise that Alloy helped us with on what the value could be. Without getting into numbers, what I can tell you is, the value in a very limited use case, purely around spoils for a small – I think two or three customers alone, forget the rest of the customers we may be working with – was probably 7 or 8X what it costs for us to be able to continue to use this platform. So the value is absolutely there.

Awesome. Hey Glenn, actually Mark has a follow-up question to that comment. If there’s no ERP integration, where does the master data come from? I think maybe add some clarity there.

I will start by, I think, highlighting Alloy’s capabilities. They will ingest information, whether it be from your retailers or whether you provide them information – we provide them with some daily transactional information, some daily, and some weekly. There are reports that get run every day in the background; anyway, we just add them to a distribution list. What they’ve been able to do, this is kind of smart when you think about it, is that there’ve been cases where I haven’t been able to provide them a unique, perfect master data solution; they sort of deduced it from the ongoing files that I send them. We send regular files about our own shipments daily, and those files happen to contain customers, customer numbers, materials, material numbers, and brand information. The following materials are all in a brand. What they do a little bit on the side is, they deduce from the information you’re sending, what your set of materials is. They can also use that as a signal to me for where I do provide master data to say, “Hey, the following 10 items are cropping up that I don’t have your full master data.”

It is a good question. I would tell you that it’s something that we work specifically with Alloy with, but it’s a capability they use based on the fact that they’re ingesting all this information in the first place, they can do a lot of deductions in the back.

To add to that, Glenn, we have a very flexible data model. So ERP data is coming into Alloy, although it is in a less direct fashion at Ferrero. That gives us some optionality in terms of whether we gonna integrate directly with MD04 in SAP, which is an option for us as a pre-built connection, or whether we’re going to piggyback on existing sorts of reporting and extracts that are already being shared. In the case here, to make it lighter touch from an IT perspective, we’re flexible in how we capture that data, whether there’s inference happening through the data model to get as close as we can, or hey, let’s actually consider bringing in different sorts of channels. So yes, ERP data absolutely is in the application for Ferrero, but with direct ERP integration, we can be flexible in how we think about capturing data there.

Yeah, I think that’s a good answer to that.

We got another question about whether you, or how you restructure your organization, to use this tool?

I don’t know that we found that we had to restructure, necessarily. Again, still a journey.

This year, I would tell you we’re looking at new user groups. If you want to think in terms of functions departments, last year was mostly supply chain – maybe I’ll call it 75% supply chain, 25% sales. It is rapidly, now it’s probably 50-50. Going into next year, there may be some folks in marketing who are looking at this, or trade marketing.

We haven’t necessarily restructured, but I think the real key is that the same information has different value to different functions for different reasons. I know that seems like a mouthful. But the point is, I wouldn’t know how to describe how to set up an alert for potential lost sales; that’s something a salesperson can work with Alloy on to design. I know how to talk about where I look at inventory imbalance within my retailers, or where I look at end-to-end inventory and understand that.

I don’t know necessarily you have to restructure, I don’t think it’s necessary, but I think it’s really key for each function that engages to speak with Alloy specifically about the kinds of things that they’re interested in looking at because their language – much like the language of different retailers – their language is going to be different from sales to supply chain, to trade marketing, to consumer research folks.

We have found that the important thing is that, and it’s actually a good thing, because the tool doesn’t have to be everything to everybody. It can be X for one set of people, Y for another set of people, but the reality is that it’s all running the same data in the background.

I know we also got a question here about what the minimum percentage of coverage of POS data is? I feel like I asked Logan that question when we were originally looking at exploring a partnership with Alloy. I now think that the answer is different than what I was looking for. Of course, I was looking for something like, well, if you don’t have 75% of your sales, maybe it’s not enough. I don’t think that’s the case. What we have found is that it depends on your use case.

The use case we talked about, about spoils, earlier. For us, there are two or three customers that drive, let’s call 75% of that. Those two or three customers, maybe are only 15 percent or less of my business, maybe only 10 percent of my business, so it’s a big difference. The reality is, with those two customers alone, in that use case, I get value out of it. That was the one I told you was a 6-8X addressable kind of risk on those two alone. You don’t really need the 80%.

Now, if you wanted to make sort of big, sweeping end-to-end choices, obviously, the more coverage the better. The simple answer is we had a certain amount of coverage on our first year, and this year we’re going to expand to three or four more retailers, which is gonna increase our coverage.

I would tell you we ran through the first year with probably, let’s call it 50% with respect to our sales, and still plenty of value, plenty, plenty of value.

Great question, though.

It sounds like we have some more questions coming through, but we’re running short on time here because we’re right on the limit. Logan, are we gonna take these down, get back to these people?

Yes, you can keep feeding in the Q&A. We want to be mindful of folks’ time here. Again, I really appreciate everyone taking time out of, I’m sure, a busy day, to have this conversation. Glenn, thank you so much. Really appreciate the insights and the presentation here.

If there are additional questions, we’re happy to stay on the line as well, but otherwise, we very much appreciate everyone joining us and have a great rest of your Wednesday.

Thank you.

About Glenn Lawse

Vice President of Supply Chain USA
Ferrero Inc.

Glenn optimizes the balance between service levels, distribution cost and inventory through continuous improvement in supply chain efficiencies at Ferrero. He leads a team across supply and demand planning, warehousing and distribution, inventory optimization, and customer service and logistics.

Glenn joined Ferrero in February 2017. Previously, Glenn spent more than 15 years within Johnson & Johnson in roles of increasing responsibility in the areas of supply chain management, supply chain strategy, manufacturing network strategy, ERP strategy, and supply chain & IT project & portfolio management.

Glenn holds an MBA in finance and supply chain, MA in Italian literature from the University of California, Los Angeles and a BA in comparative literature, Italian literature and English literature from Cornell University.

About Logan Ensign

Director, Client Solutions
Alloy

Logan is an expert in predictive analytics. At Alloy, Logan works closely with customers to help them maximize value from the data, analytics and planning platform by ensuring fast implementation, delivering trainings, sharing ongoing best practices and conducting regular business reviews.

He joined Alloy from InsideSales.com, where he led the company’s highest end service, Momentum PRIME. For customers who wanted to use predictive analytics to transform their sales operations, his team formed long-term relationships focused on optimizing sales process and strategy and ultimately delivering and showcasing value.

Logan’s early career was at RIC Insurance General Agency, where he worked in Corporate Strategy and Sales. He holds a degree in Biology and minor in Economics from Stanford University.