How to Fulfill Orders When You Don’t Have Enough Product to Meet Demand
By Yossi Sheffi
Director, MIT Center for Transportation & Logistics
The global shortage of semiconductor chips that is disrupting supply chains in the automotive and consumer electronics industries is forcing chipmakers to decide which customers’ orders to fulfill when product is in short supply.
How do chipmakers allocate product fairly, profitably, and in such a way that they do not alienate customers and inflict long-term damage on their businesses?
This problem often arises when large-scale disruptions create scarcities of vital components for an entire industry. Companies need to think about how they are going to address this challenge.
Multiple market forces
Much of the current problem is rooted in the well-known bullwhip effect in supply chains. In the first stage of the pandemic, many automotive manufacturers reduced their orders, anticipating a downturn in vehicle sales. The impact on microprocessor foundries positioned high upstream in the supply chain was exaggerated, and they experienced even larger order reductions from suppliers to the automotive companies.
Subsequently, demand revived, but at the same time, demand for home devices that use electronic components skyrocketed, and chip suppliers turned their attention to those industries. Another driver of the current shortage of chips is the US reliance on a relatively small number of chip foundries, especially the giant supplier, TSMC.
Since these drivers will not disappear any time soon, deciding which customer orders to fulfill is a problem that will continue to vex chipmakers for the foreseeable future.
Fortunately, there are tried and tested frameworks to tackle the product allocation conundrum. I describe them in my latest book The New (Ab)Normal: Reshaping Business and Supply Chain Strategy Beyond Covid-19.
Multiple factors to weigh
As the Washington Post reported on March 1, 2021, the approach some chip suppliers are currently taking is to put big and powerful customers such as Apple and Samsung at the front of the queue, leaving less influential companies to wait in line for supplies.
This is one approach to the problem – but it might not be the best. Here are some factors and strategies to consider.
Favor the most profitable customers
One approach is to direct limited supplies to the highest-margin products and customers. For example, General Motors scrambled to find scarce materials in 2011 after a trifecta of disasters—an earthquake, the resulting tsunami, and the related nuclear meltdown—hit Japan and devastated factories there. At one point, GM could not find enough airflow sensors for its trucks. The management team prioritized full-sized trucks over small trucks because the larger vehicles were both more profitable and had smaller retail inventories. Thus, GM temporarily closed a Chevrolet plant in Shreveport, Louisiana, which made the (small) Colorado pickup truck. (As it turned out, shortly before the plant closure, more parts were found, but it was too late, and it took another week for the plant to reopen.)
Favor strategic customers
This is probably the approach taken by chipmakers in the above example reported by the Washington Post. Suppliers who use this framework consider customers’ current volumes, the growth opportunities they represent, and customers’ ability to switch suppliers. Thus, some companies favor their biggest customers. During several disruptions to microelectronics suppliers over the past 25 years, the largest PC makers, including HP, Dell, and Apple, were high on many chip suppliers’ priority lists.
This approach’s downside is that a simple product profit margin calculation can ignore the long-term importance of a customer to the company.
Rationalize SKU count
When the Covid-19 crisis hit, demand for consumer staples outstripped manufacturers’ production capacities. Consumers began eating more meals at home and stocked up for sheltering in place, which drastically increased sales above historical levels. Makers of nonperishable foods had to allocate their overtaxed production capacity across various SKUs.
Many enterprises chose to suspend production of low-volume SKUs to focus on the top sellers, simplifying their supply chains by focusing on fewer products (and SKUs). General Mills, for example, cut its Progresso soup line from 90 to 50 varieties and eliminated some flavors and package sizes of breakfast cereal. Thus, they ensured their best seller availability and affected cost reductions owing to less product changeovers during manufacturing.
Fairness to all customers
Some suppliers insist on “fair” or uniform allocations of volume because of commercial, cultural, or legal reasons. With a uniform allocation policy, all products or customers get identical treatment, such as the same fraction of ordered volume or the same upper limit on the number of items (e.g., “limit two cartons of eggs”). After the 2011 Fukushima nuclear disaster, many Japanese companies gave every customer the same fraction of their orders. Likewise, as a large supplier in the PC industry, Intel generally uses a similarly uniform allocation approach to avoid the appearance of favoritism.
But fair fractional allocation is not easy when customers try to game the system by artificially inflating their orders. To combat this, some companies allocate product based on a portion of pre-disruption historical order volumes. However, in some cases, the disruption does affect actual demand, such as the supply of medical devices and PPE during the pandemic, and some suppliers have taken this into account in their allocation algorithms.
Favor vulnerable buyers
Allocation by customer vulnerability can be considered if the product is life-saving or essential to the customer’s survival. Amazon, for example, prioritized “essential products” such as food and medical supplies in allocating its limited fulfillment and shipping capacity when the Covid-19-related increases in e-commerce outstripped the retailer’s abilities to fulfill and ship orders. Similarly, some retailers catered to vulnerable customers during the pandemic, such as reserving the first opening hour for the elderly and other at-risk populations to provide favored access to freshly restocked shelves in a freshly cleaned store.
Even if the supplier’s preferred allocation method favors larger customers, it may be willing to divert small quantities of supplies to ensure the survival of a smaller enterprise customer. Verifone, a maker of credit card processing equipment, wasn’t a large buyer of the electric motors made scarce by the same 2011 floods in Thailand mentioned above, but its absolute dependence on these motors led suppliers to fulfill its (small) orders.
A Rubik’s Cube of competing interests
The product allocation strategies described above reflect trade-offs between the supplier’s self-interest to remain profitable and its customers’ interests.
However, these trade-offs will change in line with changing market conditions, so the strategies must be revisited regularly.
The key drivers of the current chip shortage will be with us for some time, and the supply issues have prompted a re-evaluation of the semiconductor market that may result in structural changes. This possibility makes it even more important that chipmakers revisit their approaches to order fulfillment regularly because the future semiconductor market may look different than the current one.