Welcome to the first podcast on Supply Chain management dedicated to the healthcare industry. This podcast is brought to you by Supply Chain Operations a boutique consulting firm based in Switzerland dedicated to helping the small and medium sized pharmaceutical company on their Supply Chain matter. Today, we are discussing our first topic, which is demand and supply forecasting, and the tool that we have developed to support that activity. I'm happy to have that podcast and to be here today with Maria Lewis data engineer, mathematician and expert in supply chain. She's the person that developed the tool and that is helping different companies to implement it. I'm also here with Gaspard former global head of Supply Chain at Novimmune, and Gaspard will be sharing with us his insight on the tool that he has implemented and the different learnings that it brought to Novimmune. So let's get started. Maria Luz can you tell us what's the tool about and the different functionalities it offers?
Yes of course. The planning tool is an Excel tool intended to evaluate requirements at all levels, finished products, drug products, API, critical material. This evaluation are based on the demand data, but also on the supply configuration and constraints. The tool will recalculate supply plans for all product levels each time the demand forecast is updated, or there is a change in the upstream manufacturing structure and constraints. The results will be given in a table format, which can be easily exported, but also in a graphical way, highlighting the various values that are normally display in the production plans, including demand, stocks, safety stock levels, and replenishment slots. I would say that the strength of the tool is to include and nlink all strategic components in one single computation, also allowing to get visibility on the capacity to optimise workload and throughputs. Another very useful feature is the possibility this tool gives to test different scenarios. At any time within the planning process, you can interact with the tool and develop scenarios based on alternative sets of assumptions, and each scenario can be saved for comparison purpose.
Excellent. We all know that scenario testing is extremely important. Can you give us a bit more insight about that and how does it actually work?
Yes. As mentioned earlier, you can easily modify the set of assumptions to evaluate the impacts on the results. The most common concerns demand assumptions. What if demand is X percent lower or Y percent higher than my baseline? But of course, you will also be able to modify other parameters in the model. For example, safety stocks, production intervals, or minimum order quantities. What we usually would recommend to do is to calculate your supply plans based on your best assumption and without any consideration of frozen period. Then you fix until some point the calculated supply plans and you stress them by increasing or diminishing your demand by playing with your safety stock levels or lead times or whatever assumption you feel may vary in the future. And this will give you very, very interesting information concerning how robust your plans are. The goal, of course, is to optimize these plans to decrease both stockout risk and write off risks. Playing with scenarios could also be very useful to do sensitivity analysis. How do the variable impact outcome? What are the most cost efficient adjustment that can be made to the design? These are examples of questions that you can address by doing scenarios.
And if I may add to that, it sounds that there's a lot of things that the tools can take to account and that might sounds like a lot to go through for a company that is just jumping into it. However, what you are not saying right now is that you are behind company that are implementing this tool and it is huge asset to have you design the tool.
Excellent. Indeed it's true Gaspard. It is very important that it's a pragmatic tool end-to-end and it's easy to implement. So we will come back on that notion a bit later. But I'd like to discuss a bit further about risks. So you refer to demand uncertainty and the different scenario testing. In your experience, can we leverage the tool for other risk management and, how does it work?
Yes, very interesting question. It is true in the supply chain when we talk about uncertainty, we first think forecast accuracy, and it is true that demand is a major uncertainty factor, but risk may come from other sources. You're right, failures during the production process, supply shortage of key components, transportation disruptions, and so on and so forth. For this more complex situation with many sources of risk that need to be assessed, I would recommend adding a layer to the tool and integrate a Monte Carlo approach. Rather than just replacing the random variables with a single, single average number, a Monte Carlo simulation will assign a probability to this variable variables. For instance, you may said that each API production slot is a random outcome with let's say 10% of probability to fail. Then you run a large number of simulation and you average the results. The Monte Gallo approach itself is quite simple to understand, very intuitive in its principle, less simple to implement, but it gives a powerful evaluation of your overall risk when you are in this situation of having many sources of risk that you want to assess in their wholeness. And again, this is a feature that we can implement in the supply tool.
Thank you so much for sharing. Before we move to more concrete elements on how it can be implemented. Could you enlighten us and share who is this tool made for and what are the companies that can most benefit from implementing such a tool?
In my opinion, the tool is particularly relevant for a company that does not have a new ERP system yet, and is moving from a development phase to a commercial phase because this company needs to secure availability of the products, taking into account all the manufacturing constraint, as well as all the various lead times prior to the first commercial launch, meaning that between 12 to 18 months before launch inventory building must take place.
I would like to add to that it's also a tools that it's preferable to go through while being thinking about having an ERP because getting into the ERP world for that, you need to have a pretty good idea about what you need, what you want, and what are all those constraints that you're talking about and defining right now. This tool helps a lot figuring out what the ERP should look like later on. I would also add that not all company, well, depending on the size of the company, would afford to have an E P that could be as good as these tools in terms of functionality. So it's probably for many company going from research and development, clinical into the commercial world, a tool that could remain for a little while in order to manage this planning activity.
Yes, absolutely right. It's kind of a of a ladder and it's a first step into planning activities before jumping into a complex ERP environment. You can learn with the tool, develop some learnings and then probably define business requirements and then move into, let's say full speed with let's say large scale model and with an ERP system or any other advanced planning tool.
Flexibility would be a, would be a term that belongs to the tool, which obviously when you get into an ERP you lose a hundred percent of flexibility. So it's very well made for getting started into the business of commercialization. But it could also serve as managing a clinical study, our clinical planning, buildup safety stock, etc. So not only for the commercial world, but even more for the commercial world.
Thank you for, for sharing your insight. I think it's very valuable to have your feedback from the ground. And I'd like to finish by asking how easy is it to implement? and maybe if you can walk us a bit through how long does it take to implement? what is the effort for the organization that is implementing the tool in terms of resources and also in terms of data?
In general, we can set up a first version of the tool in one to maximum two months. During this period, we'll organize workshops with the company to gather all the necessary data. And depending on the availability of this data and its structure, I would say that the company will have to invest maximum five person days to provide to us all the needed information. And I, I would add that it is a very interesting exercise for the company gathering all this information because sometimes we realise that it is perhaps the first time that they are putting together all these relationships that they were not necessarily aware of the impacts of all this linked data connection.
At Novimmune, did you experience the same? and can you walk us through, you know, how, how easy it was to be implemented?
Yes, sure. So that was, that was quite easy. As I was saying before, it's not a tool that just hand up on your desk and you have to deal with. Maria Luz is behind you, guiding you through. There's a lot of parameters that you may not have think about, but obviously are very useful to set up. So just having Maria Luz bringing us through was just an ideal situation. The number of things, the number of patterns that she already has, been through that she could think out of the business when you describe that to her is helping tremendously. So I would say extremely easy, extremely easy to maintain as well, because she doesn't leave you after a few hours, a few days. You may have new idea that she would pick up and improve into the tools. She would also guide you through if one day you just forgot how it works or you are not sure about what you're doing, she would help you to interpret your data. So the, the tools is already a great asset, but also having someone such as smile with her experience behind it is just a state of the art situation.
Well, thank you very much both for sharing your insights and to let us know more what is the demand and supply engine tool about. It was really, it was really nice to do this first podcast with both of you. Thank you everyone for listening. You can help us by leaving us a review or making a comment about this podcast. And we will be back about every month to share more Supply Yhain insights on the healthcare industry. Thank you for listening.
Read the article related to the Demand & Supply Engine.