Demand Planning and Forecasting
No one can predict the full scope that this pandemic will have on consumers. What organizations can do, however, is react swiftly in the short-term, prepare for what is to come in the mid-term, and plan for recovery measures in the long-term.
The rapidly growing effects and impacts on demand have varied greatly on the types of goods and services offered across the different channels. To help manage the impacts of COVID-19, or any future disrupt, on the demand planning processes, there are steps that can be taken at each phase of disruption based on approaches to past events combined with demand forecasting principles.
During the Disruption:
So far, the timing and duration of the pandemic have varied geographically, while many Western markets still remain within the disruption’s lifecycle, other global regions have begun to move toward a post-pandemic recovery stage. Depending on which stage in the lifecycle and the industry in which your organization operates, demand may be artificially high or low. For example, goods that consumers deem to be essential may be experiencing an unusual increase in demand, causing stockouts and scarcity, while other industries may be seeing a drop in demand. There will also be new shifts in behavior within organizations’ data as demand shifts from brick-and-mortar stores to direct-to-consumer channels. In response to unexpected behavior, traditional demand planning models tend to over-correct on forecasts, and often produce trends that lag actual results.
During the disruption, especially in the recovery stage, demand forecasts require adjustments. Given the lack of historical data on a disruption this large in scale, applying direct alterations to forecasting models is troublesome and time-consuming. Forecast modifications can be applied to ease the process and provide a swift approach. The key facts that forecast modifiers should include are:
- Timing: Since the timing and length of this disruption is still unknown, modifiers should consider ongoing analysis. Assumptions should be reconsidered and evaluated frequently when new data becomes available.
- What-ifs: When establishing scenarios and ongoing revisions, best and worst-case scenarios should be formulated with a ranged degree of effects.
- Channels: Online consumer activity should be monitored and evaluated frequently to determine if consumer traffic is higher or lower than a pre-pandemic period. Naturally, until brick-and-mortar stores reopen, direct-to-consumer will be the primary liquidation channel and should be leveraged to avoid inventories of unsellable items.
- Products: It is also important to analyze the types of items most affected during this demand disruption. During a crisis, some items are deemed essential by consumers. There was a rush on toilet paper causing stockout everywhere before stores shut down, but it will be just as important to anticipate potential rushes to products and services once the pandemic ends. There may be goods or services that have a pent-up demand that should be anticipated as consumers return to their normal lives.
These factors can be modeled by leveraging channel data that continues to provide demand insights and data from regions that are further along in the overall disruption.
After the Disruption:
Once the effects of the pandemic begin to subside and life returns to normal, demand data that would normally be included in forecasting models will be significantly affected. Unless additional information is included in the forecasting models, demand forecasts will be skewed by the actual data recorded during the pandemic’s disruption. Identification and application of variables that affect historical data will help ensure future demand forecasts are not heavily influenced by outlying data from this period.
It is important to note that the pandemic’s disruption will likely result in a new normal for months or years in some industries. Implementing analytical techniques is the best way to overcome the effects of this event and optimize opportunities for recovery.
Advanced Analytics Forecasting
About Vanguard Software
Vanguard Software introduced its first product for decision support analysis in 1995. Today, companies across every major industry and more than 60 countries rely on the Vanguard Predictive Planning platform. Vanguard Software is based in Cary, North Carolina.