Study from Blue Yonder and WMG takes a long look at impact of COVID-19 on retail supply chains

It is fair to say that the COVID-19 pandemic has had a jarring impact on retail supply chains in myriad ways, in terms of how retailers had to respond and react to a new way of handling their supply chain operations. That was made clear in the findings of a recent study published by Blue Yonder (formerly JDA), a provider of AI-driven and end-to-end supply chain management services, and WMG, an academic department at the University of Warwick, England that provides research, education and knowledge transfer in engineering, manufacturing and technology.

Data and feedback in the study, which was administered online by Qualtrics in late April, was based on insights from 105 senior executives in retail supply chains throughout Europe, Asia, and the Americas.

The study’s findings focused on myriad areas of retail supply chain operations, including:

• the majority (61%) of retailers used inventory to buffer against the disruption of COVID-19, adding that supply chain processes and systems were effective, but more than half (58%) of retailers said a high degree of manual intervention was required to respond to the fluctuation in demand and supply;
• workforce issues were dominant issues for retailers with 59% of warehouse and 48% store operatives being affected by quarantine or illness, which often resulted in the closure of online operations and the need to recruit temporary staff.
• retailers were polarized in their treatment of supplier payments, with 37% delaying payments and 30% making early payments;
• 40% of respondents cited an aspiration to operate fully automated warehouses in the next five years;
• 29% of retailers relied on suppliers with more agile manufacturing and distribution networks, which the study indicated is a potentially more resource efficient and resilient response;
• retailers are at an early stage, with only 12% of respondents “digitally ready”; and
• with 75%-to-80% of products seeing demand fluctuation, retailers were slightly better at responding to decreases rather than increases in demand

To get a deeper dive into the study’s findings, Logistics Management (LM) interviewed Jim Hull, Senior Director of Industry Strategy at Blue Yonder.

LM: The study noted that 58% of retailers indicated a high degree of manual intervention was required to respond to the fluctuation in demand and supply. Can you please provide some examples of that manual intervention? 

Hull: Good examples of what we saw are store order overrides – essentially a store manager starts to see that a product is selling faster than ‘normal’ and tries to get ahead of it. They create special orders for more product that gets shipped from the DC to the store, but many stores doing this causes the DC to quickly run out of product. The DC Planner then has to override their normal order quantities to try and get more product, including going into currently open purchase orders and increasing the quantities.

Of an interesting note, we did see that retailers leveraging our AI ordering engines picked up that store increase in demand faster and reacted automatically. This allowed the entire system to sense the demand shift faster and allowed those retailers to get back to the suppliers a bit earlier than other retailers, giving them ‘first bite’ at the available supply.

LM: 40% of respondents noted an aspiration to operate fully automated warehouses in the next five years. What needs to happen/occur for them to have that goal come to fruition? 

Hull: To achieve a fully automated warehouse there are several pieces that need to be put in place:

• automated order reconciliation and prioritization – essentially the Warehouse Control System needs to know how much product is available, how much product is needed and when there is not enough product available for the orders being requested, how to prioritize which orders to fulfill first;
• automated picking – this is the ‘guts’ of the automated warehouse and depends on some form of automated transport capability as well as conveyors and storage assemblies that enable both case and piece picking.  As a note, in some cases, this may include having to go back upstream to the vendor and modify the product packaging to make it ‘pickable’ by a robot; and
• in fully automated facilities, the receiving, put-away and shipping steps are also automated, using autonomous pallet trucks to move product off and onto trucks and even automated pallet building robots that optimize the stacking of product on the pallet for efficient transport and downstream handling.

LM: Why are retailers better at responding to decreases in demand rather than increases in demand?

Hull: From a planning perspective, traditional replenishment systems tend to quickly react to slowing sales, but lag when sales pick up. 

In a slowing sales environment, when the replenishment system checks available inventory vs. predicted demand, it will see that there is still plenty of inventory, and thus not create a new order, slowing down the whole supply chain relatively quickly.

When sales pick up, though, the system sees the inventory is being pulled through the system, but typically uses a moving average or similar approach to ‘guess’ how much of the recent trend to believe. Most retailers I talk to want to see a trend continuing for a few weeks before fully betting on the increased sales, so they keep their replenishment systems on the conservative side. This works fine for repeat surges like the 4th of July or Holidays where you have a sense of how high demand will go, but in the recent crisis no one knew if the spike was real or short term and therefore were slow to react.

Additionally, it is much easier to ramp down capacity in areas where it isn’t needed than it is to ramp up an already highly utilized portion of the supply chain. For example, in Grocery, the frozen foods chain depends on refrigerated storage and transport trucks. When people were suddenly eating home much more, the frozen food sales took off. However, grocers couldn’t easily add more frozen storage at the DCs, nor quickly procure a bunch of refrigerated trucks, so they struggled to meet the increased demand in these somewhat specialized areas.

LM: Why is the percentage of "digitally ready" retailers only at 12%?

Hull: In one word, “Data.”  An autonomous supply chain requires clean data, at the right level of granularity. This includes clean inventory data at all points in the supply chain, solid product information, sales information, pricing history, promotional history and digitized promotion plans, etc.  Most retailers are only now beginning the journey to create an enterprise data resource that can feed the digital supply chain.