The beauty of this forest is that we have full control of this. If you're buying a system where we're a big black box and you want to have something changed, then you need to make a service call in each FIC consultants. They'll give you an offer. This is what it's going to cost here. We have guys in OT and they can program it overnight.
And next morning, we would run it with the new algorithms. So to say so far, it actually. It brings a lot of agility in both ways, both in terms of adopting to genes conditions, but also for the plan and the day to day life. 
Good morning, good afternoon. Or good evening, wherever you are in the world. 
This is the 
digital supply chain podcast.
The number one podcast, focusing on the digitization of supply chain. 
And I am 
your host global vice president of SAP.
Hey, everyone. Welcome to the digital supply chain podcast. My name is Tom Raftery with SAP and with me on the show today, I have Phillip. Phillip, would you like to introduce yourself? 
Yes. Hello, Tom. And thank you for having me here. My name is Phillip Melchiors. I'm the, the head of the. Local planning team at Velux, a global manufacturer of roof windows.
I've been in the working world as a bison for, I think 20 years. I actually started out, uh, back at a university, getting into inventory optimization and doing all kinds of mathematical modeling in with areas. And when I was, uh, when I first my, my thesis, I was looking for a place to, to, to try that out in a, in realized.
And I just. I wrote out to a lot of companies and then that who's looking for someone like me, and then I actually entered the book 18 years ago. So I've been in the same company for that for the entire period. But when I'm working in, in, in many different areas of, um, logistics and in particular really much into inventory optimization and network optimization, My last eight years focusing really much under record the production planning parts.
So the whole S&OP process and M and a, and then also little by little what it systems can support this. And, um, Oh, what time I got into management as well. So today I'm not the specialist doing the spreadsheets myself by, but actually working as a, as the head of the department. 
Super, super congratulations for people who are unaware it.
Could you tell us a little bit about Velux. 
Yeah. So, so if he looks, it's a, it's a, it's quite a family owned company. I think we have more than 75 years. Um, and we produce mainly windows for the, for the roof. Uh, this is the whole, uh, the, the philosophies that we once would bring lights and, uh, and freshly out through the roof to, to, uh, to the customers.
So, um, it's a Danish based company. It's, uh, but with today we have, uh, sales companies all over the world and I want you to production all over the world. Uh, Europe is our. Our biggest region. So that's why it's, but also Americas is coming. I'm coming on. So it's quite well. Um, yeah, we have about 14,000 employees and a net revenue just below 3 billion Euro last year.
So, uh, it's in, uh, it's a four for guy like me. It's a, it's a, it's a quite big company. It's a lot of complexity. This is again, it's a, it's a company that's grown over time and we have, um, we have a quite. Integrated supplies in the sense that we own a lot of different processes. So we have a lot of responsibility for the products and all the way up to the finished goods.
So when you're sitting in planning like I do, that's a lot of interesting complexity and finding out how is motors and pains and wood pieces and hardware all fit together into the total product. So quite an interesting place to be. Yeah. Yeah. 
Yeah. And I mean, we're in the middle of a global pandemic. Um, I think you guys must've been impacted by this obviously changing demands and suppliers able to supply or not able to supply, or can you, can you just tell me a little bit about how COVID-19 impacted the, the, the roof window industry?
I certainly can it's it's been a really a roller coaster ride for us. I think at least sitting in planning, of course I'm there, it started in our European markets and the first Margaret's that impacted was Italy and Spain, which was, I think also the ones generally hardest hit in Europe and we saw in a matter of one week or so the sales simply dropped to zero.
In this market. So quite, quite dramatically actually, and there, and we saw all the other bigger markets was going down as well. They were not down to zero, but at the same time, governments were coming out with all kinds of, of lockdowns and, and, and, and recommendation. And we decided actually on a higher level in bills to, to, to close a lot of our factories for two weeks, it's just to, to, to predict our employees and support the local government's initiatives.
Also because we are quite honestly expecting that's two weeks production. We would have used the loose in no time because sales were dropping so rapidly, fast, but then the forecast cost of ride started because we then actually saw some markets just before locking down. We had some small holding where we say associates wind up and then the wind competed down.
So we said, okay, fair enough. People are probably more wanting to, to having some kind of, um, Products at home. They can go away. If they're forced to stay home for three weeks, at least they have something to do. So that may be explain it. But the problem was that a lot of our markets, our problem, it's a good thing for weeks, but sales didn't drop at all.
So we will be sought was a hoarding, was just the beginning of a period with increasing sales. And we were close to our life with all the marks and finding out what is going on here. Is this still just hoarding or is it, is it actually, um, increased sales or, or, and then little by little, some of the other pros came, Marcus came back from the lockdown.
So, uh, so you, sales wise have been really, really pure with a lot of uncertainty. Normally B looks quite stable. We are, we, I. We look at products, some people installing themselves a lot use installers. So somehow there's a natural, I won't call it bottleneck, but that's less capacity on how many windows you would sell.
And so we don't see big differences in sailor, which is making it very nice. And, and forecastable product
changes in this. And, um, And the problem was we had post op production. Of course, we then opened it again. Eh, but we were winning from a planning point of view where I'm responsible for securing that we have the products and stock when in this situation where our stock was record low, our supply versus demand was.
Completely unpredictable because really we're talking minus 50% and plus 50% for four big markets, which has a huge impact. Our products are relatively country-specific in the sense that they're the type and the size and so on. Uh, actually quite locally, even if it's a global brand, but that's different tastes if you like between the different countries.
So, so it's not just the same window, so we need to have the right products and stock, and then it's been really, really tense. And on top of this constant, also the fact that supplies now steeply, and might we have suppliers own internal suppliers for whatever reason, they are also impacted by this. And we have production sites, which has to close down for the full factory.
We haven't experienced that yet, but part of the factory have been closed down due to, they had a few employees with Corona and we had to then close down for two weeks. And so that's enough. Yeah, uncertainty and what is also supply, right? So it's been really, really exciting periods for us to, to be in and, 
and, and I hope you responded.
As good as we can never say. We are generally a company that takes a lot of pride in our deliberate forms. Our ability to deliver on time to our customer center. We have to say we haven't been able to live up to the same standards as we normally do, but we'd be afraid. Maybe if we try everything we can.
For us, it's been, eh, normally an August a little bit more into the area of planning. Normally we have a monthly it's Opie process, so that means every month preview review the sales forecast for the coming period. And we have an executive meeting where we sit together with sales, supply and finance. And then logistics and say, okay, what is the outlook for the computer?
And we take the bigger decisions in our, in our planning to reduce or increase production and whole our it infrastructure is set up to, to, to, to, to, to, to this ministry process. And right now it's a, B, it's not normally what happens in the Monteith happening there, eh, all on, in, uh, in this period. So we simply had to, to increase the frequency of these meetings.
So we ended up having this. Coming into a weekly frequency instead. And then we were really also our dialogue with the markets, the way we organize this, that we have a central organization that is responsible for the old forecast when we get a lot of input from the sales companies and that we had to also.
Increase the frequency of, of course. So some magnets, we even had two weekly meetings to simply understand what is going on in each market. Because eh, if first of all, a lot of things, when we will live with the police, just different governments deciding what kind of law. Down, eh, we will, you will, you do that changes very rapidly, but also then understanding what is the impact of this?
In some countries, installers are allowed to go out to our customers and all the windows and, and some, they are not, and it's a huge difference for us. Right. And understanding this is actually, that's something that's, it's hard. And even if you had the conversation and the input from local guys, we were on quite a lot still, but we're better off than not having that dialogue.
Of course. So we, we increase the, the, the, the frequency of the product, uh, of, of the whole, of the whole process to, to a weekly one. And for us, that meant that we were able to, eh, what we produce next week is more close to actually what it's going to be. So in general, our products are stocked. So that means that we do have some buffer in the supply chain, but still when the stocks are really, really low and sales are high and not according to forecast.
We easily end up with the wrong products and stock. So this was our way of minimizing. That impact is so that that's been, um, this, this, this has been our response and I think so far, we're doing well right now, sales are still really, really high. And we are also the point where our production see that this is, this is our maximum limit.
Uh, we also had to implement different safety messages in our production syncing, and then also. Cleaning out between different shifts. And now also for, for you to say for business continuity, ensuring that there's no overlap or shifts. So if sickness break out is only once, if you have to take out, but the fruit factory, so, so capacity is limited, more limited normally.
So, so we, at a point right now where we have said, okay, let's, uh, let's, uh, Pull all leavers and, um, and this is what we will do. And, and then it's up to us to balance as good as possible. And so far we are fine, but it's a long summer we have in front of us. And, and there'll also be some vacation and we are working with Saturday sifts and they have reduced their vacation, but it's, um, it's going to interesting.
Yeah. Cause I I've, I've had conversations with a number of organizations throughout this. And it's interesting because I don't know if this is the case with Velux, but some manufacturers, you know, their lines were not set up for physical distancing between employees. And so for them to maintain the same kind of throughput, they either have to stand up new lines or they have to increase the amount of hours that are producing are they have to increase automation.
Otherwise their productivity goes down. So is that some kind of trade off that you've had to make as well? Or were you already set up for physical distancing? 
It was actually a, I think there's been a few cases where we had to modify it, but it's been quite limited. We had the two periods, two weeks closure, but where we were, we were using that two to two weeks to two to make these adjustments to, to be able to produce.
So. So I think in general, we have been quite lucky. I would have to say that the way we have designed our lives lines is makes this possible. Right. But no one would know that in advance, right? Yeah. 
And then where, where from here? I mean, we, we have, you know, possibly another 18 months of this kind of variability and then what does kind of post Corona, uh, supply chain look like?
Uh, what, what, what do you see going forward? 
Yeah, it's interesting because I think what we can see now is a little bit going back to the financial crisis 10 years ago, or even more now, or that suddenly we are in a position where the world is much more on unknown, right. And uncertain. Uh, right now we see positive sees and I think we have.
We hope, of course it's just a 10 year growth, but realistically, we also think this is an outcome of people not going on vacation. They're staying at home for the summer and they'll probably invest in a, in their house. They spent two, three months inside and working at their home office and it's been a, it's been probably two or three, one months inside.
So some daylight and fresh is probably, it's probably a good time for that, but that will probably not last right. Once you're back in, in a normal. Way of life, then you will start traveling again and you'll not stay at home. And, and, uh, when we do roof, so realistically our sales go back to more normal level.
We don't know when it will happen, but so for now, it's we end up in emergency a moot. Planning wise, but sales wise, we of course enjoying it, but, uh, our, uh, but from a supply chain point of view, looking forward, I think we will start looking into that. Yes, it's very stable period supply wise demand wise, but the future seems to be that we will be fluent a big lesson here that, uh, prepare for the unforeseen as well.
Right. Because you, and you think this will live with that. It did. And it's very easy to imagine that something similar can happen. For sure. It's just so easy having three or four years where things were unstable and you forget the past, I think this has been quite a lecture, so I think it will take a longer time before we I'll be sort of forgetting the past year.
Yeah. Yeah. And I mean, if we look at some of the technologies that are coming down the line and the kind of AI and machine learning and blockchain and all these kinds of 
things, 
do you see those being big advantages for people like yourself? Do you see yourself rolling out any of those technologies or how do you see kind of future technologies, helping companies like yourself?
It's a, it's a, it's interesting, because of course we're looking into these things and we have areas in Middlesbrough where we are utilizing this and making full benefit of it. But actually in our planning system, it's relatively limited how much we have implemented. And I think that's, um, that's, I think different for what all the people is doing because they start to, like, we don't see the potential in that.
But, but there's also for us, we have a long supplies and we control them the production of wood. We control proximate pains. And then finally, they're simply if the windows and you have a whole product of accessories around it. And our problem is we have a very, very highly automated planning system. Then there's a lot of these factors that need to be.
In the system as well. So, so in our case, when we do a production planning, if there's a part missing somewhere, then the system needs to know otherwise they will make the wrong plan. So we've actually gone a little bit, the other way around and said, let's not make it a fully automated planning here, a system here.
So, so we, uh, we have made a planning system where we actually have Xcel as a, as a front end. It is if a P is IPP the new, a new, a new planning platform that you're offering. But what were you really making use of? It's not all defensive functionality where we are making use of it's an Excel front end.
And that means that from the system, we will get a proposal for a plan, but for us, and I think the foster strengths in list slice, and that we have an Excel front end where the planners can easily just go in and type an eight it's the plan. And it force it's really. Getting the specific details from the facts to say, well, this Thursday, we only got to run at 50% capacity because one machinery is of maintenance or whatever, the reasons.
And that means that then we can go in easily modify this and save it back to the system. So for us, the most important part is actually not to have the automated part, but to, to combine it somehow. So you get a good proposal from the system, but it didn't have to do that. So you have the flexibility to go in and make the, the manual work item.
It would be perfect if you could have full automation and close the lights, of course. But, but, but just for us, there's a really long way. And it depends on the stability of your production. Again, the predictability on what is actually, what is life going to be like? And the forest, when we have 14,000 employees and a lot of them isn't manufacturing, that's also a lot that can go wrong.
It's a physical process and realistically things. It doesn't always go according to plan and we need to adopt that into our systems. Interesting. It's a, 
it's a very kind of contrarian attitude where everyone is saying move away from Excel. No, no, no more Excel. And you're going the opposite way and saying, no, we want to use Excel because I could just more flexibility.
Yeah. It's actually, we have quite a few discussions on that in the management group was whether or not that was the right day. Way to go. And of course there's some, there's some, some disadvantaged sex they'll that people go in all kinds of different directions, but the way we are using is a P and IBP, it's like, it's done in a way where you still get a lot of the standardization in the tool.
So it's not just. Well, each fan of having their own spreadsheet and the two of the plan, the plan has done IBP. We just get the output out on an Excel front end and allows us to, to, to do the digging in there. We also add some, some functionality, which is based on the visual basic actually really, really old programming stuff.
But that's what is coming with a right. So, so if we can be a free program, macros that will do standardize things, cleaning up stuff, leveling out our production plants. Taking out for historic house, making small adjustments if it's allowed and so on, all kinds of things. And the beauty of this for us is if we have full control of this, if you're buying a system where we're a big black box and you want to have something changed, then you need to make it.
So it was called in each FIC consultants. They'll give you an offer. This is what's going to cost here. We have guys in OT, they can program it overnight. And next morning we would run it with the new algorithms. So to say, so for us, it actually, it brings a lot of agility in both ways, both in terms of addicts, To, to, to change conditions, but also for the plan and the day to day life cost, we try to, we want to make sure that still, that, that things just don't go out of the system.
So, so that's a lot of things we do to control people, stay in, in, in the XL version of IBP and not just go in a separate sheet and so on, but a lot of it is going. A little bit in the other direction then than you would normally hear in the industry. Yes, that's correct. Yeah. 
Well, I guess, I mean, it has the advantage.
Everyone knows how to use Excel 
East panel, their favorites will, everybody will end, right? Because it's so, had so much flexibility and you can easily just copy paste a number of, of, of, of sales and make a graph or a summarize that, or what do whatever you want. You can then. Multiply probably about 10% and put it back.
That's all of this functionality is just embedded in the tool and people in know it from a, for 20 years. So, so they are more or less fluent in, in the, in, in the planning tool from day one. Right. So it's, it's a huge advantage when you start out with, at least 
we're at now Phillip, the 20 minute Mark. So we're, we're about to wrap up.
Is, is there anything that I haven't asked you or anything we haven't touched on that you think would be interesting for people to be aware of? 
Oh, I think we've been quite well around a time. I don't have any here. Any other things on my mind except to saying that yes, we are working with Excel, but also I think for us, one of the next steps is going into the demand sensing part.
I think there's some, some options there where machine learning can really help us. I'm sure. In a situation where like the coverage. It's probably is going to be hard for even machine learning to, to adopt to that. But we also expect a more normal sales and force. This would be the area where we will be looking into what can, how can we get this into it?
Because the short term forecasting, I think we have still a, we can, we can, we can, we can, we can improve a lot. I still an a and we have a lot of benefits that have to be harvested. I think this is where, at least in our situation, that, that the more modern technologies. Not just Excel sheets can actually make a big difference for us.
So this is, this is some of the areas we will review and look into in the coming ones. Super, super 
Phillip. That's been fantastic. Thanks a million for coming on the show today. 
You were able to come. Okay. We've come to the end of the show. 
Thanks everyone for listening. If you'd like to know more about digital supply chains, head on over to sap.com/digital supply chain, 
or 
simply drop me an email to Tom daughter after he had sap.com.
If you'd like to show, please, don't 
forget to subscribe to it in your podcast, application and choice to get new episodes. As soon as they're published 
also, 
please 
don't forget to rate and review the podcast. It really does help new people to find the show. 
Thanks catch you all next time.