The digital supply chain is making progress. However, significant potential for value creation remains untapped. In this post, we reveal the success factors for digitizing end-to-end supply chain planning and creating new business opportunities.
As we explained in our last post, the digital supply chain is out of the starting blocks, with lots of exciting ventures and achievements to show for it. However, we have captured only a fraction of the value potential so far. Digitizing and automating end-to-end supply chain planning is the big prize, but many companies struggle to put all the pieces of the puzzle together. Digitization of end-to-end planning needs to bring many different parts together in the right way: changed processes, new roles and skillsets, clear decision rules, the proper templates and workflows, enabling data as well as supporting IT and data infrastructure. From our work with leading global companies in supply chain planning, we know that three success factors typically separate the wheat from the chaff when it comes to digitizing end-to-end supply chain planning:
- Data – intelligence – execution: Start with a focus on data quality as an enabler for a digital supply chain and establish a common data model. Based on that, disaggregate planning into distinct, manageable decision loops, which transform data into optimal planning decisions (in Systems of Intelligence) and automatically feed the results back into execution systems (Systems of Record). Each of these loops should help make a distinct supply chain decision (e.g., segmentation, allocation of scarce product, or adjustment of parameters) more precise, automated, and value-adding for the company.
- Slicing the elephant: Approach these decision loops one by one with a clear target roadmap in mind, step by step. Optimizing one planning challenge at a time in a decoupled approach (e.g., first focus on automated inventory buffer sizing, then on allocation and ATP) helps create tangible value early on while not overwhelming the organization and creating undue implementation risk on the business side as well as the IT side. As digital planning maturity grows, companies can then move to a more integrated approach, where different optimizations are balanced from an end-to-end perspective.
- Focus on the machine-to-human interface: Focus not only on the technical improvement potential (e.g., optimizing algorithms, using new machine learning techniques), but also zoom in on the human improvement potential. Effective decision support tools supporting planners, e.g., in the plan review or data validation step, can often have significantly higher value contributions than sophisticated new algorithms.
What are the key digital use cases in end-to-end supply chain planning which companies should consider first and foremost in their digital strategy? From our perspective, the 10 cases outlined in figure 3 should be top of the list. In the following section, we will explain two of the key value opportunities in more detail.
Figure 3: Top 10 digital value opportunities in E2E supply chain planning
Smart inventory & availability alerts. Machine learning pattern recognition has the potential to fundamentally enhance forward-looking visibility and alerts regarding shortages or excess inventory. Traditional approaches track one metric at a point in time (e.g., actual vs. target stock). Smart alerting algorithms screen multiple time series, for example, stock, orders, demand, and forecast. These algorithms correlate the time series, track their rate of change, and apply machine learning pattern recognition. By doing so, these algorithms can detect over- or under-stocks with a much higher probability, thus supporting the supply chain planners with fewer and more high-quality alerts and improving overall supply chain reliability and service level. Adaptive, dynamic buffer adjustment. Digital and analytics can also help make the supply chain more adaptive by flexibly adjusting replenishment parameters (e.g., reorder points, safety stocks, and delivery quantities) to the current demand and supply situation and its risk and variability profile. Machine learning algorithms can use data like demand, demand variance, and supply variance to automatically adjust the parameters mentioned above on a daily basis. This approach helps planners implement a parameter-driven planning approach where tactical planning adjusts flexibly with suitable risk buffers. Consequently, this approach allows manufacturing operations on the shop-floor level to perform more stably and reliably, without avoidable rescheduling and firefighting. The below case of a global appliances company illustrates the benefits of this approach.
Case Study: Adaptive Buffer Management at a Global Appliances Company
The company faced strongly seasonal demand, poor forecast quality on the SKU/location level, and fragmented, decentralized planning processes. These numbers illustrate part of the complexity: Planning dealt with over 100 stocking locations and more than 50,000 SKUs in the end-to-end supply chain. Inventory planning and execution were identified as key focus areas of the end-to-end transformation and the company implemented adaptive, dynamic buffer management which used demand and supply data like forecast variability, orders, and bottlenecks to make daily adjustments to key replenishment parameters such as relevant forecast horizons, reorder points, and safety stock levels. This helped both stabilize production planning and improve overall service and inventory levels. In summary, the performance improvement was substantial: The customer service level increased by four percentage points, while inventory decreased by 20%, and planning efficiency rose by 20%.
Companies need to think strategically about digitizing their supply chains to be able to secure significant value and competitive advantage. In particular, they should begin by focusing on the following lead questions:
- Which supply chain decisions have high potential to be enhanced with digital and analytics?
- Where can we generate the highest value for the business while the change effort for people and technology is still manageable?
- Where do we find use cases that both improve supply chain performance and customer/user experience?
- How do we slice the elephant and build a logical roadmap with clear and early value cases but without overwhelming the organization and creating an excessive burden of IT implementation?
What is your approach to digitizing your supply chain? Let us know your thoughts and feedback! More insights in part 1: Race for Value: Towards the Digital Supply Chain