André Martin, the father of Flowcasting and DRP, and Fred Baumann, JDA GVP of the Flowcasting Business Unit, have presented the principles, advantages and early customer success stories of the ground-breaking Flowcasting process during several well-received webcasts. Supply Chain Nation asked Baumann to provide a brief summary of the learnings and what a Flowcasting-powered world might look like.
SCN: You and André covered a lot of ground in your webcasts. For the benefit of those who were not able to view those yet, could you please provide a brief summary of the key take-aways?
Baumann: There a few key take-aways that I would highlight. The first one is about how we must change our thinking. You have to unlearn the old ways of doing things and some of the traditional supply chain planning practices that are dominant today when you enter the more connected world of Flowcasting. What does that mean? First, there is transitioning the planning model itself. Many manufacturers are still basing their consensus demand plans and their S&OP reviews on orders and shipments out of their distribution centers and plants. The Flowcasting model moves to a connected model where you are focusing on the sell-through to the consumer as the basis for the forecast rather than the sell-in to the retailer.
The second unlearning, or really the extension of an existing planning model, would be the inclusion of your trading partner’s network in your planning process and systems. Many manufacturers today within their supply chain forecasting, planning and distribution plan limit the scope to the edge of their network with their finished goods DC. And for retailers, the edge of their finished goods DCs is the end of their planning model. What we are talking about here is a joint plan that is inclusive of both trading partners’ networks from the shelf back to the factory and enabling the parties to collaborate on that common plan across the network.
The last unlearning centers on the fact that there has historically been collaboration on the outputs from two independently generated plans. CPFR actually started this thinking a few years ago, but now we have this common model of the business from the shelf back in a common system. Therefore, we can collaborate not just on the outputs of two plans, but rather collaborate on the inputs of a single plan. The inputs are things that drive the flow of goods, such as promotions, supply chain parameters, or the network itself in terms of the way product is flowed from the plant to the store shelf.
The second major highlight I would point out is one of the key principals that we reinforced over and over again—that there is only one place you should forecast and that is at the point of consumption. Andre makes a great point that you should never forecast what you can calculate. There are too many people forecasting up and down the supply chain. But if we have a really good handle on the consumption forecast at the shelf, we can calculate up the supply chain and eliminate a lot of redundant forecasting, whether that is at a manufacturer’s DC, a plant or a retailer’s DC. We can start with the shelf-level forecast and net the forecast up for inventory, lead times and all of the other supply chain parameters that cause orders to be generated, and the rest is a calculation. That was a key principal that was highlighted.
The next take-away I would highlight is that Flowcasting can dramatically improve parallel processes. If I have a common plan of demand, and how that demand is translated into orders up each node of the supply chain, it becomes a great input to parallel planning processes such as transportation capacity planning, warehouse planning for loading and unloading at the distribution center, labor requirements for nodes in the supply chain, financial planning and S&OP, and the SKU plan at the shelf. For example, we can model flow plans based on anticipated changes in the planogram, including planned changes in safety stocks and presentation stocks. Having a time-phased forecast at the shelf enables planners to have space allocation plans that incorporate future demand patterns and seasonality.
The key point in one of the webinars was that in order to gain critical mass in Flowcasting, you need to be technically capable. Manufacturers historically planned their supply chains at the finished goods DC level. Extending this to the shelf results in millions of SKU-location combinations. You need a solution that can scale technically based on volumes, and just as importantly, scale from a planning perspective because you need to be able to take on these larger planning intersections without significantly growing your planning staff. You have to have a process that supports planning by exception and uses intelligent analytics to focus on the biggest issues and learnings that help you understand the root cause of issues at the shelf to save planner time and mitigate risk that the issue will occur again in the future.
Another webinar highlight is that in order to create a reliable time-phased forecast from the shelf back, you have to have a process that can support slow and intermittent demand. Sixty to seventy percent of a retailer’s items in many cases will move off the shelf at an average rate of one unit or less per week per store. Many retailers have therefore taken a min-max approach – “one to show, one to go.” To get the real benefits of Flowcasting, you must be able to translate those slow-moving items into time-phased plans accurately. JDA provided some insights on how we support that. We also spoke about the JDA Cloud and how there are new options to help retailers and manufacturers scale very quickly and economically in this type of relationship.
The last thing I would highlight is that the size of the prize for executing Flowcasting is huge. Having a common plan across the entire supply chain that is visible and jointly executed by both parties takes a lot of risk, buffer and waste out of the supply chain. Companies that have taken this approach have consistently seen significant improvements in on-shelf availability, and they have been able to do so with less inventory. They have also seen lower distribution and transportation costs through improved load planning and efficient network flow planning.
Those are the high points I would cover about the webcasts.
SCN: In our next post in this Flowcasting series, Baumann will discuss what the supply chain world would look like after widespread Flowcasting adoption.