The Franke Story: Three Points to Learn about Supply Chain Complexity
Nowadays everyone’s writing and speaking about supply chain complexity, but not many are talking about how to turn around all that complexity and variability for business advantage. That’s exactly what Franke, a world leading provider of solutions and equipment for residential kitchens and bathrooms, was able to do.
According to Inbound Logistics, there are three main sources of supply chain complexity: meeting the needs of customers, globalization, and internal pressures. Franke was facing all three. The company’s tremendous growth resulted in a very complex supply chain network that was no longer running optimally. Franke’s Kitchen Systems and Water Systems divisions consisted of 42 locations with four echelons serving 146 markets and more than 125,000 SKUs (equating to 1.4 million SKU-Markets). Seventeen different ERP systems were in use, with varying data structures and planning processes. While Franke had standardized on SAP as the master database for all locations where SAP was the ERP system, its homegrown demand forecasting and planning tool hadn’t changed and had many shortcomings. This posed a huge problem given the complexity of Franke’s supply chain network. Fortunately, Franke was able to tame that complexity partnering with ToolsGroup, and following are three quick points to learn from their story.
Point to remember #1: Average forecasting just isn’t good enough for today’s complex supply chains.
Franke wanted a solution that could improve forecast outcomes fast in a multi-echelon environment. It also needed to support inventory optimization and replenishment and factor in high-level production constraints and varying calendars through the network. ToolsGroup’s probability-based forecasting was a huge factor in their decision. “Other tools only let us create an average forecast which was not ideal given Franke’s diverse and complex product portfolio,” explains Enrico Casalino, Head of Logistics Planning & Engineering at Franke Kitchen Systems. “Also, others required us to manually pick different algorithms for different tasks. This was impossible given our 1.4 million SKU/market combinations. Only [ToolsGroup] SO99’s algorithm decides automatically what best matches our needs — that really made a difference.”
You can think of probability-based forecasting as an “uncertainty decoder” because it identifies a range of demand outcomes and the probability of each of those outcomes occurring, in order to calculate optimal inventory targets. Probability forecasting is much more than a statistical method. It allows you to consistently place better inventory bets than your competitors, especially for harder-to-forecast items.
Point to remember #2: Adopting digital supply chain planning doesn’t have to take years…or an army of planners.
Franke didn’t have to wait long for results. The first forecast was available after only three months. At the global level, two people manage the project–one dedicating 50 percent and the other 70 percent of their time. According to Casalino, “We have a very complex supply chain network and only two people from the global team to work on the project.”
Apart from ToolsGroup, all other solutions required 2-3X the effort and resources that were at our disposal. No one actually thought it would be possible to manage this complexity with just two people.”
Point to remember #3: Easy collaboration is a must for consensus forecasting.
Contrary to some beliefs, demand collaboration doesn’t mean simply sharing the responsibility for a bad forecast! Participation from a large team doesn’t have to be complicated. At Franke at the local level a team of 100 planners across the organization can now collaborate on the demand planning and forecasting process through the ToolsGroup Demand Collaboration Hub (DCH). In this process Service Optimizer 99+ (SO99+) provides the baseline forecast to DCH. Franke planners work in parallel to add their market knowledge to the forecasting process. DCH then generates statistics that weight different data sources according to the level of accuracy they contribute. Sources that historically contribute to high forecast accuracy are weighted more heavily – and vice versa.
“ToolsGroup’s DCH approach gives us a level of visibility into the planning and forecasting process that was impossible with our old tool,” says Casalino. “The local people can contribute with their market know-how but the control and final sign-off remains with the global team.”
Most importantly for Franke, ToolsGroup has helped the company tame extreme complexity to quickly reach its forecasting targets set in the beginning of the project. The company receives reliable forecasts from all its markets, is able to distribute the forecast efficiently via the consensus forecasting platform through two echelons, and is now confident to significantly improve inventory optimization and replenishment. Read the full Franke case study below.