One aspect of COVID-19 that companies are beginning to tackle is demand planning and forecasting. Whether a company’s demand fell out entirely for a couple months or a company was selling more than ever, we are seeing inconsistent demand patterns year-over-year due to the impacts of COVID-19. From this arises a very critical question, how do companies account for the heavy variation in demand when looking into their forecasting methodology?
First and foremost, we know how important customer input can be to the demand planning process, but this is going to take center-stage as businesses begin to see COVID-19 numbers decline and the world return to a somewhat normal state. Customer insight will be absolutely paramount in the demand planning process as we do not necessarily know whether these very low or very high data points will continue into the future. Most likely, they will not, which means statistical forecasting models are going to be over and understated depending on the circumstances. Having the customer collaboratively work with sales and demand planning teams will greatly reduce the risks of inaccurate forecasting under normal circumstances, but will be even more useful as demand planners look for guidance to curb future variation in demand.
The next part of this is history manipulation. I know, history manipulation or data manipulation never sounds like a good thing, but we very well may need to consider looking at the data and trying to reconcile the history to reflect what may have happened without the global pandemic. There are clear risks here. Perhaps a customer wants and needs have permanently changed as a direct result on the pandemic and adjusting your history to try to improve models will cause an even greater error versus actual sales. There can also be a loss of information as turnover in the company would result in a loss of knowledge if meticulous notes were not taken to depict what actions were taken to adjust in the need to revert back to the actual figures
***Here we can see an example of how COVID-19 impacted variation in demand. The pre-pandemic sales average hovered right around 14,600 units. During the pandemic, the monthly average shifted to roughly 3,800 units.
I personally believe that the most likely outcome will be a shift in the consensus process. Forecasts may be simply a baseline number for the coming months where market intelligence from the field will play a more significant role in moving the forecasts up or down. While adjusting history is an option, no one knows the aftermath of the pandemic and what exactly it will do to demand along with the continuing supply chain challenges companies are facing. By using true historical values to provide the baseline forecasts, supply chain managers and planners will need to be in constant communication with the sales teams on deviations from the plan. This will lead to an overall reduction in silos as these groups work more collaboratively. It should also be noted that there may still be some true variability in demand for the coming months for a number of reasons. Among those reasons, companies or customers trying to “catch up” from their ongoing supply chain challenges and secure stable supply from their upstream supply chain.
How about you? What are you discussing with your teams in how to approach the wide variability in demand you are seeing?