At the onset of COVID-19, out-of-stocks plagued retailers. Their immense impact was amplified by the "Bullwhip Effect" to create tsunami-like waves in demand from retailers to brands and on to their suppliers. The vicious cycle of hoarding and resulting out-of-stocks fueling further hoarding forced brands to look beyond point-of-sale (POS) data as their ultimate demand signal.
Over the last 20-years, consumer goods companies and their suppliers have been on a quest to incorporate the holy grail of demand signals into their forecasts: POS data. A billion-dollar industry comprised of giants (SAP, IBM, Oracle, etc.) and best-of-breed competitors emerged to help brand manufacturers harness POS data in demand signal repositories to improve demand forecasting. And yet, the dual treasures of eliminating out-of-stocks while minimizing inventory remain elusive.
A primary obstacle is the refusal to share POS data by many of the fastest-growing retailers. Entire retail channels such as e-commerce and convenience stores are virtual black holes. How are a brand’s demand forecasters to react when consumer purchases shift from POS-visible retail channels to those hidden behind a veil of secrecy? When sales suddenly disappear from POS-visible retailers, what percentage of it is truly gone vs. hiding out at a neighboring retailer?
Another challenge in using POS data to forecast is that sales spikes result in stock-outs that prevent further sales. This leaves forecasters trying to guess what demand would have been had the product been available to purchase. No thanks to the behavior of our fellow hoarders, POS data indicates large increases in demand for breakfast cereals, granola bars, soups and frozen pizzas. Yet the brand managers and forecasters of these products don’t know the proportion shelved away in pantries vs. being actively consumed at a higher rate. This makes it impossible to predict ongoing elevated demand vs. a one-time blip.
To further complicate matters, as consumers' preferred brands run out-of-stock, they're showing a willingness to try new brands at unprecedented rates. Will these trial purchases lead to repeat buying that fundamentally shifts demand from market share leaders to their emerging competitors?
Another factor weighing on even the most prescient of forecasters’ minds is the question of shifting consumer sentiment and its potential to impact demand for particular goods. How are consumers who are fearful of a major economic depression already changing their purchasing behavior? How can brands use these shifts as an early warning signal for what may come as more (or fewer) consumers share those economic concerns?
Finally, even the most sophisticated approaches to forecasting are ill-equipped to predict how 30 million consumers will respond to losing their jobs. No model is capable of trending the effects of such a sudden shock to the system because there's absolutely no precedent for it in modern history.
All is not lost, however. Leading consumer goods companies have already risen to each of these new forecasting challenges by taking another giant step downstream — into the world of consumer purchase panels. Leading panels continuously track the item-level purchase behaviors of households and often trigger surveys to consumers based upon their purchases, both in-store and online.
Unlike typical demand forecasting improvement initiatives that require large investments tied up in multiyear initiatives, many brands have already implemented panel-powered best practices in response to COVID-19. For example, every consumer insights professional surveyed at Numerator’s recent customer advisory board meeting reported that they started supporting their demand forecasting colleagues with panel-based insights in April. When asked if they would continue to support their demand forecasting colleagues once retail conditions return to normal, their response was unanimous: “This is the new normal.”
A member of the board of directors of Numerator, Jared Schrieber is a recognized expert in the field of retail data analytics after shaping how over 500 consumer goods companies leverage retail point-of-sale data. Prior to co-founding InfoScout (now Numerator), Jared led product, professional services, customer operations and business development for Retail Solutions.
A member of the Board of Directors of Numerator, Jared is a recognized expert in the field of retail data analytics after shaping how over 500 consumer goods companies leverage retail point-of-sale data. Prior to co-founding InfoScout (now Numerator), Jared led product, professional services, customer operations, & business development for Retail Solutions.