Solving Today's Supply Vs. Demand Quandary using machine learning
The current health crisis has upended 2020 and brought challenges that many could not have even dreamed about earlier this year. For grocery retail, the challenges have been especially taxing as consumers have changed their normal, weekly shopping patterns in favor of stockpiling certain items or utilizing grocery delivery and buy online, pickup in-store (BOPIS) options.
With shoppers embracing new avenues, retailers have had to shift their technology and practices to keep up with the unprecedented change in behavior. From new operating procedures across all facets including labor, buying, forecasting and merchandising to the impact felt at the bottom line, retail margins have sent shock waves through the industry and erased the best practices that have been built over the years
For retail CFOs, reflecting on the past six months has likely led to the realization that strategic adaptions must be made to respond quickly and offset increased costs, while still providing the service that the consumer expects.
RETAILERS ARE CARRYING TOO MUCH INVENTORY
The typical weekly shopping trip that most consumers embraced has been completely overturned. While some have canceled trips to the store all together in favor of grocery delivery or BOPIS, others have continued their shopping trips but return to the store less frequently and often make bigger purchases while doing so. Seven months into the health crisis, we’re still seeing shortages and stockouts of staple items like paper goods, cleaning products and hygienic items, leading consumers to stockpile items when they are available.
These new shopper habits have continued to put immense pressure on the supply chain. Retailers have had to buy additional inventory, forcing them to carry the cost on balance sheets longer, to avoid shortages where possible. With grocery margins traditionally built on quick turns, grocers are often selling goods at retail prices before paying their bill to the wholesale supplier. With the shift to pre-buying eroding traditional margins, grocers have been forced to pay for stock well before it is sold or even delivered and carry that cost until goods reach the customer’s cart, potentially months later. This has all culminated in the fact that financing the supply chain has never been more challenging.
STOCKOUT ISSUES CAN LEAD TO CUSTOMER CHURN
Many retailers are still trying to understand how much inventory is needed. The stockout issues trickle down, often resulting in unhappy consumers, and potentially further eroding margins and causing customer churn. When shoppers don’t find their preferred brand, they will select a different option or go online to try and find their exact preference. If the shopper opts to buy the store brand instead, this can have a mixed impact for the retailer by offering a higher margin while preserving the category strain.
Ultimately, the unmet demand from category stockouts represents not just a missed opportunity for retailers, but a much bigger loss: shoppers who find empty aisles may turn elsewhere not just for out-of-stock items but for their full carts. This eliminates potential revenue and could lead to these customers never returning to the store, resulting in immeasurable lost future revenues.
THE MAGIC BULLET OF MACHINE LEARNING
With the multitude of changes over the last six months, many CFOs are left wondering what the future of the industry might look like. While inventory planning usually incorporates learnings from past years, 2020 has left many businesses without a roadmap to follow.
One change that will be a beneficial advantage for years to come is the wider adoption of machine learning (ML) for inventory optimization. Many retailers have implemented ML that simulates and forecasts potential inventory strategies and then uses the results to identify optimal inventory situations and parameters, eliminating the need for manual trial and error. This allows retailers to respond in real-time to issues and challenges that arise and adapt their strategy to allow for maximum supply chain efficiency, better forecasting and allocation.
ML can also assist with the rise in popularity of services like BOPIS. From helping to identify busy pickup times, product stockouts, employee staffing issues and more, ML can assist retailers to not only provide a better BOPIS service, but ultimately a better experience for all shoppers by freeing up resources. Utilizing technology such as license plate recognition when cars pull up, multi-order picking and real-time inventory updates allow for a faster, easier and more enhanced experience for consumers.
Six months from now, it’s harder than ever for CFOs to predict what the grocery industry, much less the world will look like. By utilizing ML, CFOs can be better prepared to manage cash flow, determine the impact of inventory on the balance sheet and ultimately protect revenues. While we can’t predict the future, CFOs can be better prepared for tomorrow by utilizing the learnings and insights of today.
