2020 has been a particularly tumultuous year, much of the economic volatility is due to the impact of COVID-19. According to the Brookings Institute, the U.S. recorded its steepest quarterly drop in economic output on record in the second quarter of 2020, with a decrease of 9.1 percent, when GDP had never experienced a drop greater than 3 percent on record. Retail sales alone dropped 8.7 percent between February and March 2020 (Brookings Institute, 2020).
This begs the question: “What will happen during the holidays?”
Rest assured, there is a path forward.
During particularly volatile times, there are things you can do to combat a lack of clarity. Three things you can do now include regularly checking the FRED Economic Database (Federal Reserve Bank of St. Louis Economic Research) for trends relevant to your industry, increasing the granularity of your analysis, and regularly engaging with company leadership.
The FRED offers a robust repository of information from Gross Domestic Product (GDP) data to Consumer Price Index (CPI) data to unemployment rate data. Important when looking at this type of data is to keep in mind leading vs. lagging vs. coincident indicators as they relate to the economy. Leading indicators are those that point toward future events, whereas lagging indicators confirm a precedent. Coincident indicators, on the other hand, clarify the current state.
Bringing it back to retail, before the pandemic, the Lipstick Indicator postulated that rising sales of lipstick are an indicator of troubled times – this was coined by the chairman of the board of Estee Lauder to describe the boom in cosmetic sales during the economic downturn in the early 2000s. Given the current state of the pandemic, many argue that is not true as people are wearing masks in public (NPR, 2020).
In terms of lagging indicators, think of the aforementioned CPI – this indicator operates on the notion that price increases can have a significant impact on the economy. It is reported monthly. Then, a coincident indicator might be something like GDP – it is a procyclical indicator that moves in the same direction as the economy.
A particularly important indicator to pay attention to in retail is Advance Retail Sales, a leading indicator that provides an early pulse on retail trade activity, which took a dip between February and June 2020. This can be seen in the transactional data of many retailers, given the dip retailers saw at till.
As one can see, economic indicators can be incredibly informative but still require careful examination as they are still just indicators.
Beyond that, in doing any Forecasting this holiday season, one ought to also consider context. Senior executives at many companies have access to aggregate information that can show them macro trends of their business. They are also likely interfacing with other key executives, maybe even across industries. They can inform you about novel events, strategy changes and even new product development that might help modulate your forecasts. While including this context in a quantitative model is an art in and of itself, when done correctly, it can provide a level of lucidity.
Finally, you should heavily consider increasing the granularity of your forecast. Considering your product hierarchy, are you forecasting at the family-level, category-level, or class-level? Why? What about the item-level (the most granular level)? As microtrends are often a part of macrotrends, having a pulse on changing consumer behavior can be incredibly informative. If your category is down, are there a handful of products that are doing particularly well or, as they say in economics, countercyclical, moving in the opposite direction?
While there is no crystal ball or silver bullet, there are strategies and frameworks that can be used to help guide the incorporation of the right variables (perhaps many of those cited above) into your forecasting tool(s). This, in principle, is the intersection between management and your data science department.
Coupling the knowledge of the trends that you typically see during the holidays with the knowledge of what might be happening on a microlevel as well as with various leading and lagging economic indicators can be a strategy for success. At the end of the day, any good forecast generated by an algorithm is best when modulated by business rules or contextual variables.
To learn more about how we generate robust forecasts during uncertain times, you can reach out to the ADC Data Science and AI team. Over the past 30 years, through good times and bad, retailers have relied on us for direction in Fresh.
About the Author
Michael N. Colella is the Chief Data Scientist at ADC where he is responsible for providing the data science perspective on everything we do and infusing our solutions with Artificial Intelligence. Prior to joining the ADC team, Michael was a Director of Data Science at dunnhumby, a global leader in customer-centric retail analytics. At dunnhumby, Michael was responsible for leading analysis for some of the biggest names in global retail, focusing on innovative initiatives. Before his time at dunnhumby, Michael was a Senior Manager of Advanced Analytics at Kraft Heinz where he led the global advanced analytics innovation hub under the CIO. Michael has worked/consulted for various other retailers & Consumer Product Goods companies, including Bayer Consumer Care as well as Wilton Brands, to name a few. Michael has 10 years of experience working on innovative analytic projects spanning Supply Chain Planning & Optimization, FMCG, Big Data, predictive and prescriptive analytics as well as behavioral and cognitive neuroscience.