In the previous article, we started to cover the root cause need for safety stock, variability. If there is no variability in the system, then there is no need for safety stock, because the numbers perfectly calculate. But we know that there are highly variable components in the real system that we model in our inventory management systems. There are those pesky and unpredictable consumers who never follow a perfect pattern. Then there are those vendors that never ship on time. Add in the crappy performance of our own company and the job of the inventory planner is almost impossible.
Oh, the humanity.
So we buy insurance in the form of safety stock, often investing in twice as much inventory for what we consume per cycle because the executives want to be in stock 98% of the time. At the same time the EVPs of Stores and Merchandising are screaming for higher service levels, the EVP of finance, Mr. CFO, is telling us we have to trim inventory. Then there are those smartass guys in the warehouse that keep talking about taking inventory by looking at the dust on the boxes of stuff sitting in storage. They just don’t get that we have to have that safety stock.
Oh, the carnage.
Squeezed into a seemingly no-win, lose-lose situation, inventory planners continue to dream about ways to get better forecast models to improve the error rate. If only we could get a better model, better demand signal data, we can do a better job of lowering forecast errors. Better tools and better data would make the job easier and we'd get some peace and quiet from the executive wing.
Oh, just wake up.
Getting better resources is just one way to solve the problem. Forecasting only addresses part of the picture. I can forecast sales, but if the humans that I sell to don’t show up, I might site on too much inventory. Moreover, if the humans that run the vendor operations, and the humans that run our distribution centers, or the humans that put stuff on our shelves screw up in executing the supply side of the supply chain, my perfect forecasts don’t matter.
Previously, I outlined five different components that affect the success of any inventory management process.
Imagine this list in a different way:
On one side of the lever are the planning activities, Forecasting and Policy. Execution activities of Lead Time and Execution Performance sit on the other side of the lever. The Resources of people, systems and capital used in the process are the fulcrum. The inventory system relies on the strength and placement of the fulcrum to maintain balance. If the Forecast and Policy are weak, the resource effort must move closer to the Lead Time and Execution side of the lever to maintain balance. If Forecasts are strong and the policy is good, the fulcrum effort moves closer to the planning side of the lever.
Planning and Execution share and contribute to the resources. In the last article, we talked about some of the resources that the planning activities need. Let’s pick it back up by looking at the resources that both sides of the lever depend on.
Cash is a resource. That is why the CFO focuses on working capital. Working capital tied up in safety stock inventory is cash that is not creating more cash. The CFO wants that inventory to turn over, to sell. Selling inventory converts the working capital back into cash. If the inventory is turning, it is generating the Operating Cash Flow that the CFO is looking for to fuel the business.
Turn the inventory and you get the CFO off your back.
Turn the inventory, and demonstrate ways you can improve the turns, and you are in a negotiating position with the CFO to get funds, cash, that you can use to get more resources, like the systems to do a better job forecasting.
There is just so far the inventory managers can push the planning process to maintain balance. The Execution side of the lever, Lead-Time and Execution, have as much influence on the balance of the system.
Supply chain managers put a huge emphasis on Communications and Visibility as the primary forms of execution control. While vital and important, they are just tools, resources to use to help all of the different parties on the platform dance the inventory-balancing dance.
The communication processes documents we use in trade (the Order, the Order Confirmation, the Shipping Notice, the Invoice) all provide visibility of the transaction process, by documenting and communicating the details of the events in the transaction.
The Control Tower school of thinking embraces the notion that a single point of control can monitor all of the activities of the supply chain and initiate changes real time in the process. While this notion of a central control tower works in some complex systems like airports and air traffic control, the scale of control is limited to what is in the range of vision of the tower. Just as the tower at Dulles International Airport does not have control over operations at Atlanta, Houston, or LAX, supply chain control towers have limits of what their organization directly controls.
A cupply chain is distributed across links in a chain. Those links each represent a different business entity, each with different goals, pressures, resources and abilities. Unlike a real chain, supply chains are weak, depending on the relative strength of the weakest link of that chain.
Ever push a chain? It is hard to do without lots of guides controlling where the chain goes. Chain guides in machines provide a disciplined path for the chain to follow when under slack, or when pushed. Like real chains, supply chains are hard to push without an extensive use of agreements for controlling factors. This is hard enough to do when all of the related supply chain functions reside in one organization. It is almost impossible to do when there are multiple entities in different companies.
Let’s look at the challenge of the air travel system. Passengers choose the airline because of price, schedule and service. Multiple airlines operate through any country, operating a multiple airports. Each airline must coordinate the equipment, flight crew, cabin crew, catering, fuel, maintenance, and flight plan for every flight. The air traffic control system manages the progress of thousands of flights at any moment, communicating with the pilots controlling the aircraft. Control passes from one traffic controller to another controller in a single control center, then from one center to another, as the plane moves across the country. This system is vastly complex, where at any point there are over 1 million people in the air. There is no single point of control of the air traffic system, because there is no way a single point of control could manage all of the activities.
Consider our visual model again. Clearly, cash is a resource. Personnel in all parts of our inventory system, in our company, at the vendors, and the service providers between, are resources. Information management & communications systems are resources. Trucks, warehouses, ships, planes, containers are all resources. No single entity owns all of these resources, nor do they control the resources. For it all to work, everybody must communicate, inform, agree to terms, and pay the other for the goods and services to flow through the system.
The fulcrum in our model is the shared resources in the process. If we are going to balance the natural imbalance in all of the moving parts of our inventory system, we have to move the resource fulcrum closer to the work.
In the end, how our combined and shared resources work together, affects the amount of safety stock we must carry in our networks. It is not just the presence of the resource but also the quality, constraints, flexibility and cooperation of the resources that affect the balance of the inventory system.