The COVID-19 pandemic accelerated the consumer shift to e-commerce in almost every industry, including food. More and more people are ordering daily groceries with food delivery services, ranging from meal prep kits being delivered through last-mile delivery carriers, to online orders being dropped off at doorsteps by remotely operated carts and inside garages by delivery personnel. But while other retailers are navigating the shifting world of reverse logistics and how customers should return online orders that don’t work out, the food industry has been handling returns for years in the form of recalls.

While food recalls are complex supply chain challenges with high public health stakes, the methods often vary without clear answers as to why, leaving consumers at risk and managers without much guidance on how to effectively handle them. JBL authors Kaitlin D. Wowak of the University of Notre Dame, Christopher W. Craighead of the University of Tennessee-Knoxville, and David J. Ketchen Jr. and Brian L. Connelly of Auburn University examine why the food recall process varies so greatly in their newly published paper, “Food for thought: Recalls and outcomes.”

Most research looking into recalls examine organizational issues and consequences, such as business structures that are more likely to result in leadership issuing recalls or their hefty financial implications. But little is known about the actual process itself or why it varies in practice, especially given how important identifying and removing potentially dangerously tainted products is to overall consumer well-being.

First, companies diagnose a potential recall situation by identifying the problem, which products were affected, and where they are located. Next, companies have to coordinate, communicating with the relevant people (supply chain partners, retailers, end consumers) about the problem at hand and remove any tainted products. The outcomes of a food recall are both tangible, such as financial ramifications and consumer harm, and intangible, such as brand damage. Uncertainty in a recall situation stems from the complexity of the supply chain, often containing many upstream and downstream members. To combat the uncertainty, managers often use a variety of embedded options, such as allowing a recall to expand over time (scale), postponing a recall until a different date (deferring), and abandoning a recall entirely (abandonment). The entire process is called recall layering, or the dividing of a recall into segments that adapt based on the circumstances of the problem at hand.

Companies are more likely to conduct an efficient and focused recall process, one that minimizes harmful outcomes, when they can quickly and accurately diagnose a potential recall. Such a process requires effective communication and product extraction, both of which are interrelated: recall communication enables the proficient extraction of a product, which in turn will minimize harmful outcomes and keep consumers safe. The more complexity added by upstream and downstream supply chain partners, the more uncertainty exists within a diagnosis and coordination, making both the recall process seem daunting and the value of option embedding rise.

Managers employ recall options as they are needed. For example, the use of the scale recall option allows a company to begin communicating with supply chain partners about a potential contamination while continuing to diagnose the issue, giving companies some time to accurately identify and remedy the problem. If, during a scale option, it’s discovered there was an inaccurate diagnosis, then managers would deploy the abandonment option – abandoning a recall entirely rather than pulling products back from distributors and retailers that’s uncontaminated.

Deferral options enhance the food recall diagnosis as well but doesn’t allow for ongoing communication between supply chain partners. Managers might employ a deferral and later, a scale or abandonment option, depending on the findings of the ongoing diagnosis.

Ultimately, a company’s ability to manage this segmented recall process, recall layering, through options shapes the results of the recall: if companies can properly manage these segments, recalls are likely to be effectively executed and harm minimized, but the inability to management them could lead to worsened financial ramifications, harm to consumers, and damage to the brand.

What does all of this mean for managers? The model provides insight into high-risk areas that companies should focus on in order to address recall uncertainty and limit variability, while also showing which recall options could be used in different situations to help decision makers manage the recall. But it also offers an understanding that could apply to other aspects of food business, such as food fraud. The deliberate contamination of product demands a different approach than remedying a product being accidentally tainted through manufacturing or transportation issues, and the insights presented could help companies understand the underlying process variations.