00:00:00 Introduction to Lokad and its mission
00:00:59 Roles and goals in supply chain improvement
00:02:59 Decision-making and automation
00:05:42 Impact of supply chain decisions on product lifecycle
00:08:10 Role of a supply chain scientist
00:10:56 Cost of being out of stock
00:13:30 Translating business understanding into equations
00:15:50 Pit crew analogy for supply chain optimization
Summary
In an interview, Conor Doherty of Lokad speaks with Simon Schalit, COO, about the crucial role of a Supply Chain Scientist at Lokad. Schalit explains that these scientists are not just data experts but also specialists in supply chain management, responsible for optimizing and automating decisions related to inventory and pricing. They build algorithms to ensure efficient purchasing and dispatch, aiming for maximum return on investment. Unlike typical data scientists, supply chain scientists immerse themselves in understanding business processes and strategies, translating them into optimization algorithms. This role involves data analysis, client interaction, and business acumen, making it essential for driving efficiency and profitability in businesses.
Full Transcript
Conor Doherty: Welcome to Lokad. When people ask me what does Lokad do, I respond quite simply: we help you make better financial decisions. That, of course, leads to the following questions: where do the decisions come from and how are they generated? Today’s guest, Simon Schalit, is COO and head of supply chain science at Lokad, and he sat down and explained to me the critical role the supply chain scientist plays in generating the decisions our clients use to optimize their supply chains. As always, if you like what you hear, subscribe to the YouTube channel, like this video, and follow us on LinkedIn. And with that, I give you today’s conversation with Simon Schalit.
Simon Schalit: I’m the COO of Lokad, which in practice means that I head the supply chain scientist team. The supply chain scientists are the ones responsible for implementing and maintaining in production our solutions for our clients. Whatever the vertical or industry they’re working in, they are both data and supply chain specialists. So, it’s a team of engineers who make your supply chain better.
Conor Doherty: All right, well Simon, thank you. When you say “make your supply chain better,” one of the reasons why I want you here is because you’re very good at explaining in concrete terms. So, when you say the supply chain scientists make the client’s supply chain better, in concrete terms, what does that mean? What are we improving exactly?
Simon Schalit: Well, the goal of decisions in supply chains is to ensure that everything, inventory or pricing, is set in the most optimized manner possible. So, if you’re talking about purchasing items for a supply chain, for example, you want the items that you’re going to buy to be purchased at the right place, at the right time, sent to the right place, at the right time, to be available for service in whatever form, depending on the industry.
So, when we’re talking about making supply chains better, we’re talking about taking better decisions. Usually, that comes with both optimizing those decisions from a financial perspective and automating the decision process because the number of decisions that supply chains for big companies have to make on a daily basis usually goes way beyond what humans can deal with. Even if they do, in practice, they cannot guarantee that it’s going to be in any way optimized.
Conor Doherty: For example, if you’re talking about, let’s say, a retail company, they might have 15,000 products in the catalog, they might have 200 stores. On a daily basis, the supply chain scientist on that account is responsible for telling clients what?
Simon Schalit: The supply chain scientist is the one who’s going to build the logic that’s going to automate those decisions. In the case that you’ve just presented, on a daily basis, the company needs to decide how many of each product to buy and where to send them. So basically, it’s purchasing and dispatch if we’re taking this simple example. In this case, the supply chain scientist will crunch the data, of course not by himself, but with the algorithms and the tools, the computer that he or she has at their disposal.
They will build the financial logic that will take the decision, make sure that whenever you purchase an additional unit to put in stock, which in fact means making a bet, this bet is optimized so that the return on investment is the best it can be considering the amount of information that is available. And last but not least, he or she will make sure that the solution automates this decision process so that those decisions can be generated in a coherent and stable manner on a daily basis for the company.
The important element is that the number of decisions is gigantic and it should not be limited just to what you purchase. It’s also what you’re not purchasing. A decision of not buying in itself is a decision. So, the scale of the decisions, the number of decisions that have to be made on a daily basis, can be quite huge.
Conor Doherty: Thank you, and that’s something I remember Joannes Vermorel, CEO, described before. Even once you’ve made the decision, for example, make it trivially simple: I’ve bought one unit. The decision doesn’t end there because even once you have the unit, you can choose to continue to carry it, allocate it, return it, liquidate it, discount it, or bundle it with something else. All of these represent financial decisions, choices regarding resources.
Simon Schalit: Definitely. The supply chain decisions are going to affect the life of the product all through its life cycle, from sourcing from a particular supplier or manufacturing it to actual distribution to the client if you’re talking about an item that you’re going to sell, or consumption or usage if you’re talking about a maintenance industry or manufacturing industry.
During this life cycle, there are going to be numerous decisions. We talked about sourcing, purchasing, dispatch, whether you’re going to use it or not, allocations. There are going to be pricing decisions, which are not necessarily viewed as supply chain decisions in general, but from our perspective, it’s a decision that’s going to affect the cycle of stock. This is definitely something that you want to take into account and optimize within the supply chain context.
All those decisions need to be taken on a daily basis, a huge number of them. You don’t want them to be taken independently from one another because they are going to have a huge impact on one another. The most obvious one is you can’t dispatch something you don’t have, or you can’t change the price on something you don’t have.
But a more subtle link can be the more you purchase, the more aggressive you are in terms of the service level that you want to reach by placing big orders in purchasing, the more likely it is that you’re going to have to tweak pricing potentially at the end of the season if we’re talking about the fashion industry to get rid of the stock that was there to make absolutely sure that there would be no shortage. But of course, the consequence is you were not absolutely sure you were going to sell at least at the base price.
Conor Doherty: Thank you. Before we get into the main topic of today, can you, in your own terms, separate for me the difference between a data scientist and a supply chain scientist? Because again, in context, when I present Lokad at a trade show or a conference, sometimes when I’m describing a supply chain scientist, they say, “Oh, it’s like a data scientist.” In your opinion, how do the roles separate?
Simon Schalit: Well, a supply chain scientist is, of course, partly a data scientist. Data scientist usually refers to a data specialist that’s going to use statistics to extract relevant information from the data. The problem that usually comes with having a team of pure data scientists is that they tend to work only with the data that is available. More often than not, it creates this sort of Ivory Tower effect where the reality that is accessible to data scientists is only the reality that is represented in the data.
In our experience, if you only look at the data that is available and the data as it is when we start the project, you’re going to miss a big part of the picture. You’re going to miss a lot of the reality of the day-to-day processes that are usually not as well documented as they should be. You’re going to miss most likely part of the meaning of the data you have access to because the data itself is not as well documented as it should be and probably not from a correct perspective. It can be documented from an IT perspective but not necessarily documented from a business perspective.
And last but certainly not least, what you’re going to miss is all the data that exists in people’s heads. Unfortunately, this is pretty important because usually that’s where the strategy of the company resides. We talked just before, we said that when we want to optimize, we want to optimize from a financial perspective. Optimizing from a financial perspective relies heavily on an understanding of the company strategy.
Trying to say, “I want to reach a certain service level,” there is no optimized service level. There is no level of service where I could say, “Oh, this company needs to have 98% service level.” That doesn’t exist. The choice of this target service level needs to rely on what the company thinks this service level is worth financially speaking.
For this particular question, which we face on a daily basis with our clients, the key element becomes what’s the cost of being out of stock? If you’re talking about being out of stock for an aeronautics MRO company repairing aircrafts, the cost of being out of stock is gigantic because it can literally mean having an aircraft stuck on the ground, which costs hundreds of thousands of dollars a day.
Conor Doherty: And the supply chain scientist investigates all of this and relays it to the client?
Simon Schalit: Yes, you absolutely need to investigate that because that’s the element that’s going to guarantee that the system, the algorithm, punishes the potential of out of stock in the correct way with the correct magnitude so that it will take the decisions, the bets of whether you want to have a particular unit in stock or not, in the correct way. For MRO activity, the service level that you want to reach is extremely high because in the odd chance that you don’t have the part you need, the cost is going to be gigantic.
On the other hand, there are activities where being out of stock is a lot less problematic because clients might expect you to be out of stock, for example, at the end of the day for fresh fruit products.
Conor Doherty: For example, fresh fruit products can be substituted with something else that you have on display.
Simon Schalit: Exactly, there can be substitutes, equivalents, or just the fact that it’s not always a problem to be out of stock. You’re not necessarily going to immediately lose clients by being out of stock. The externalities are not as huge.
So the main problem with having a traditional data scientist team is that they might be oblivious to that kind of thing. To ensure that our team at Lokad, the Supply Chain Scientists team, doesn’t fall into that trap, we purposely called them Supply Chain Scientists. This makes absolutely sure that everybody, including themselves, understands that part of their job, and in fact, a very significant part of their job, is to understand processes, understand the company, understand the financial strategy, and translate all this into equations.
I was going to say words because you need to document it, but ultimately, equations in the joint procedure manual. Definitely, you need to document that for our sake, for Lokad’s sake, but also for the client’s sake. So you document all that and ultimately translate that into mathematical terms so that it goes directly into the equations that are going to be fed to the computers doing optimization.
Conor Doherty: Thank you. If I were to summarize, the Supply Chain Scientist role is multifaceted. It’s not just working on crunching the numbers, crunching the data, using computers. As you said, there is a face-to-face interactive element where the client and Supply Chain Scientists are in regular contact to discuss the intricacies, strategies, goals, desires, and constraints.
They get all of that information that may or may not be reflected in the data so that it can be transformed into the deliverable, which, very simply put, if I’ve understood correctly, is better financial decisions.
Simon Schalit: Yes, exactly. In my opinion, that’s what makes the role of Supply Chain Scientist interesting because you have this multifaceted data aspect, the human and business aspects, as well as, of course, the statistical aspects of the problem.
Conor Doherty: This is one of the reasons I’m delighted to have you on because this is very much how I see Lokad as well. It’s like using the analogy of betting. For me, when people ask me at a supply chain event what Lokad does, I talk about decisions. How do they do that? I don’t talk about maths, computers, internet, and algorithms. I talk about the Supply Chain Scientist who is the expert. It’s like if you’re buying a car and you get a world-class mechanic who will help you. The car is the decisions or the algorithm that generates the decisions, and the Supply Chain Scientist is your personal mechanic who can fix things if something goes wrong.
Simon Schalit: I like this image. I would even go further. I would say it’s your whole pit crew if you were to talk in Formula 1 terms. It can be multiple people, but it goes beyond just repairing your car. When you talk about a mechanic, people think it goes a bit beyond that.
It goes to the point where they will choose the type of car that you need, the type of engine that you’re going to need, how this engine needs to be fine-tuned, what kind of brakes you’re going to need, and the type of tires that are necessary for the type of environment you’re going to be in.
So if you want to summarize it, the Supply Chain Scientist is your whole pit crew. I think you could see it like that, and in that way, you would understand how this crew is important for you to be able to navigate whatever environment you’re going to face at the wheel of your car.
