Today’s volatility demands intelligent supply chains

From the panic buying of toilet paper early in the pandemic, to the worldwide computer chip shortage reducing PlayStation 5s to gold dust, recent years have reminded retailers of the need for intelligent, highly adaptive supply chains.
 
It would be easy to blame COVID-19 for the turmoil and expect logistical challenges to settle again as we exit the pandemic. But more than two-thirds of supply chain executives say they’ve been battling high-impact business disruptions non-stop now for over three years. Often they have had little time to recover from one disruption before the next one hits. We need look no further than Eastern Europe for a reminder of how quickly a geopolitical scenario can escalate.
 
It seems that these challenges aren’t going away soon. Chip shortages, for instance, are predicted to persist into 2023. To safely navigate the volatile future of retail, managers must urgently invest time and resources into more resilient, agile, data-driven supply chains to help them minimise business impact and meet consumer demand.

Digitising supply chain forecasts

Forecasting demand for products and services has long been key to creating quality customer experiences and generating profit. So, it may be surprising to learn that only 8% of retailers globally implemented real-time automated processes when planning and forecasting their stock demand in 2020. This suggests a lack of digital readiness among the majority of brands.
 
Short-term, high-frequency forecasts require a customised and intelligent approach to data and analytics, leveraging a combination of structured and unstructured big data to predict and optimise supply chain processes. Demand can unexpectedly change based on a whole range of factors, be those environmental changes, shortages, marketing activity, and more.
 
Simply, the accuracy of these projections can be the difference between a store ordering the right number of barbecue supplies ahead of a heatwave, or being sold out of these products and seeing competitors hoover up the profits. Get this wrong, and retailers may see a prolonged downturn in sales, customer satisfaction, retention, and reputation.

Harnessing machine learning

The integration of machine learning (ML) into a supply chain allows data to improve these predictions of demand over time. Implemented effectively, this results in more efficient stocking, inventory availability, dynamic pricing and, in the case of perishable products, a reduction in retail wastage.
 
Nicolas Vandeput, supply chain expert and author of ‘Data Science for Supply Chain Forecasting’, describes how ML is different to conventional frameworks: “Whereas a traditional statistical model will use a predefined relationship to forecast the demand, a machine learning algorithm will not assume a priori a particular relationship (like seasonality or a linear trend); it will learn these patterns directly from a training dataset (i.e. historical demand). Then, it will apply what works best to each product.”
 
New product anomaly-detection models also help retailers spot whether spikes in demand are pertinent or coincidental. Using data and ML, these models can instantly notify managers of anomalies, such as the unexpected run on toilet paper. Then, once the retailer is aware of a widespread issue, it can trigger control mechanisms such as setting customer purchase limits on high-demand items and ordering additional stock.
 
Research from McKinsey shows that ML-powered supply-chain management has enabled early adopters to improve logistics costs by 15%, inventory levels by 35%, and service levels by 65% compared with slower-moving competitors. It’s no wonder Gartner predicts that by 2026, more than half of supply chain organisations will use ML to augment their decision-making capabilities.
 
But predicting short-term, high-frequency changes in demand is only the beginning. How can retailers then go on and meet them?

Solving the operational challenge

Using insights gleaned from new ML capabilities, businesses must create more nimble and accessible supply chains to meet fluctuating demand. They must also ensure that their warehouses and shelves are stocked appropriately for expected, and unexpected, volumes of orders.
 
Part of the solution to this challenge is to make distribution centres more localised. Once you know what will be needed and where, you can plan accordingly, ensuring items are close to this demand and fulfilling orders more effectively, reducing the time they take to reach consumers. One approach, identified by McKinsey in Germany, is using stores close to urban centres as delivery hubs for rapid fulfilment.
 
Decentralising a distribution network can also broaden a retailer’s reach, outpacing competition from HQ-based challengers and meeting modern customer expectations for ultra-fast service. Plus, closer supply chain proximity reduces the chance of disruption – and can even accelerate net-zero advancements through reduced carbon emissions.

Forecasting retail’s future

Predictive analytics and intelligent supply chains are no longer just ‘nice-to-haves’. Amazon’s frictionless retail experience with one-day or same-day deliveries has set new standards in the sector and customers now expect retailers to meet their demands at a moment’s notice.
 
Delivering on these expectations requires visibility across supply chains. Retailers need to know what inventory they have and should be able to give customers a realistic timeline for fulfilment. Even if Amazon-like same-day delivery isn’t possible, the most important aspect is that retailers communicate the status of orders effectively and accurately with their customers.
 
The ability to track deliveries or check in-store stock online has become a powerful part of the customer experience, with UPS data showing 96% of customers actively follow their purchases’ journeys. This is an essential part of the customer experience: psychologists claim that the anticipation of a product can be even more exciting than the product itself.
 
To meet modern demands, customised data analytics solutions for supply chain design and optimisation are key. In the face of ongoing volatility, businesses must ensure that they have the right infrastructure and solutions in place to ensure their customers receive what they want, when they want it.

---