Everyone is seeing the dramatic and widely varying changes in retail operations due to the COVID-19 pandemic. But the biggest challenge facing them by far is how to forecast demand for the “new normal” as it serves as the foundation for almost all aspects of managing inventory and margin.
Some believe the best alternative for retailers for the foreseeable future is to abandon their forecasting systems and revert to manual methods, given that traditional forecasting systems rely almost entirely on historical sales. However, the limitations of manual methods, such as the inability to assess and forecast tens of thousands of rows of data, is exactly why we cannot go back in time — especially since the forecasting mantra to successfully navigate the current situation is “near-term and forward-looking.”
Being able to simulate multiple scenarios, as quickly as possible, will be the key to successfully “resetting” the company to the future economic and consumption state. This starts with having an AI-based, agile forecasting platform that can be fine-tuned in near real time.
Right now, the most forward-looking retailers are using non-traditional and external data sources to build new forecasting models. This includes, for example, utilizing online sales performance during the store closure period, public data on state/county/city re-openings, and demand trends from regions as they re-open to adjust historical demand for the forecasting of future demand.
Others are simulating demand using markdowns and discounting, first to determine how much budget will be needed in sell-through existing merchandise, then continuing those simulations weekly based on the latest sales data to anticipate demand trends as stores re-open.
Regardless, these companies understand that humans alone are not going to be the solution and are doubling down now to ensure their success as they come through this unprecedented event. And if we see a second wave to the pandemic, or see additional disruptions that cause other shifts in demand that have not yet been experienced, such as import limitations or product shortages, they are already well positioned to continue working in this unique state for many more months.
It's widely accepted that the farther retailers and suppliers forecast into the future, the more it will return closer to normal. In the interim, companies need to prepare so they have the right data and tools to model scenarios for the rest of 2020, into 2021 and then beyond.
More specifically, they need to use this time to learn how the changes in the economy, price and assortment, as well as the percentage of online versus in-store, will impact future shopper demand to adjust as practical to the “new normal.” This is the time to look at the whole demand process and move to being consumer-focused, not internal supply chain focused.
As someone far wiser than me once said: “When others are slowing down and waiting to see what happens, winners step up and make the most of the situation.”
Craig Silverman is CEO of Antuit.ai.