Competitiveness and performance are crucial for any organization. The quest for improving these factors is a never ending journey, and one of the key enablers is –the ability to do more with less. In this series of posts, I will share my experience about how to do more with less, in different parts of the organization, this third post deals with Inventory Management in Supply Chains.
Doing more with Less within supply chains means having a greater availability with less inventory, and yes it is possible to have close to perfect availability with much less inventory.
Inventory inaccuracy is expresses by two key phenomena – Shortages and Surpluses. The higher the inaccuracy of the inventory the more prevalent are these phenomena. Therefore, let’s try to understand the issue of inventory inaccuracy.
What affects the inaccuracy of the inventory? Two key factors; The forecast horizon and the sample size. In the previous post I focused on the forecast horizon and the simple way to meaningfully reduce it, thus significantly improve forecast accuracy and as a direct result the inventory accuracy.
In this post I will focus on the issue of sample size. When we forecast consumption we do that based on historical data of consumption. What data of consumption is commonly being used for that purpose in the supply chain. Well, to start with, all participants of the supply chain are preparing their own forecast, based on their own data. So, maybe a better question is – which of the forecasts is the key driver for the supply chain operations? Here it becomes simpler to respond. As, mostly, the trigger for supply chain operations are orders, it is clear that the forecast that drives the operations is the forecast done by the ordering entity in the supply chain. And by that logic, it is not just any ordering entity, but explicitly the last node of the supply chain, the one closest to the “end consumer” of the supply chain.
As you can imagine a supply chain having the shape of the letter “V” the entities that trigger the operations are the ones at the top of this “V”. If you consider the concept of sample size, any entity at the top of the “V” can see only a small fraction of the overall population this “V” is servicing. Thus, using for it’s forecast only the consumption information of that population.
In statistics, the smaller the population used for forecasting, the greater the error of the forecast. Thus, it is common in supply chains, that the least accurate forecasts are used as the key driver for the supply chain operations.
It makes much more sense to make the forecast, at the bottom of the “V” as there we can see the whole population being serviced and thus get the most accurate forecast possible. This has meaningful ramifications on the way a supply chain needs to be operated as well, so that the more accurate forecast also drives a meaningful improvement in the supply chain operations.
If the most accurate place to forecast is at the bottom of the “V” wouldn’t it make sense to also hold the inventory there? And when the inventory is held at the bottom of the “V” not only what s being held is meaningfully closer to what we need, but also as the supply time for most of the entities in the supply chain, turns to be only transportation time, we also can: hold much less inventory across the whole supply chain and respond much faster to changes in demand. Combine this with the concept of daily ordering discussed in the previous post and what you get is a true demand driven supply chain where with meaningfully less inventories you can experience meaningfully higher availability and reduce your supply chain operating cost. A great and simple way to achieve much more with much less.