The supply chain demand planning process in most companies, regardless of industry, will not come out of the COVID-19 period unscathed.
This global disruption might have a silver lining: if managed appropriately, the crisis can bring improvements to the planning process. This first quarter in the global economy has rendered the history unusable for forecasting. Supply chain planning must graduate to a real planning process where assumptions, economic indicators and even forecasts are used as datapoints to develop supply chain and business plans that best position the company to meet the likely future.
Before this coronavirus, the demand planning process in many companies was too heavily based on forecasts. The forecast may have been a sales team’s view of future demand, a result of statistical calculations or perhaps even a forecast provided by customers. The main hurdle for planning was the forecast error. The go/no-go factor for having a formal supply chain planning process was a forecast that would be deemed “accurate enough.”
Forecasts have always had a degree of error, and they always will. Gartner Hierarchy of Metrics benchmarking research shows that the median forecast error across industries ranged from 24-37% (see figure). Improving forecast accuracy is a good thing, but it should not be the primary purpose of demand planning and the main objective for demand planners.
As the business world closes the books on a very difficult first quarter, companies still must plan for the future. It is vitally important that every company has a plan (or plans) for the remainder of the COVID-19 disruption. Even more important is planning for the time after. Eventually, markets will settle to some levels of new normal, and the companies that have built plans and playbooks for it are likely to come out of this ahead of the pack.
The end of COVID-19 is completely unknown — nobody knows when it will be over or what the global economy will look like. I don’t know if we’ll even know when to say it’s over. There are no forecasting models or algorithms to give us an accurate projection to guide us forward with certainty. At times of significant disruption, any forecast accuracy a company may have had will go out the window.
Gartner has recently held a series of special interest group discussions that hosted a total of 50 sales executives. These leaders reported that the predictive models they have used in the past have been discarded. They acknowledge that the normal business physics simply do not apply in this environment. These chief sales officers (CSOs) report that taking a “hypothesis-led” approach with their customers has resulted in excellent collaboration. Their story shows that the demand planning process will have to go through a mindset shift, from reliance on calculated demand predictions toward probabilistic, educated assumptions-based planning and collaboration.
In the post-COVID-19 world, planning will have to be assumptions-based for quite some time. Best-in-class demand planning processes have always included an assumptions base; perhaps this will become an established component of a standard planning process.
When external market conditions are outside of the company’s control and unknowable, the plans must be built upon assumptions. An assumption is not a mere random guess of SKU volumes. A good assumption, or set of assumptions, are top-level business descriptors that are educated, clear and measurable. For example, data from Chinese economy’s recovery in 2003 from the SARS epidemic shows that retail sales growth returned to pre-epidemic levels in a matter of a few months. Assumption that similar recovery could follow the COVID-19 epidemic may be quite reasonable. It would meet the characteristics of a good assumption because it is quantifiable and it can be tracked and measured.
High-level assumptions can be disaggregated with adequate confidence to detailed lower level to drive supply chain execution. A process like sales and operation execution (S&OE) is used to track and monitor the execution of the plan. Many companies have increased S&OE meeting frequency from weekly to two or three times a week for the sake of responsive decision making. When assumptions-based plan execution is monitored, the actual demand and/or supply rates can prove or disprove the assumption. If the assumption is proven false, it can be refined or reset. This is not a catastrophe by any means — it is better for the business to have been operating on a common plan that was not fully accurate, rather than not have had a plan at all.
Business disruptions like COVID-19 are extremely difficult and stressful for everyone. However, the old proverb says that necessity is the mother of invention. Undoubtedly, many companies will come out of this difficult time with a whole new appreciation for collaboration and remarkable process improvements.
Perhaps the supply chain planners and the business decision-makers’ comfort-zone will end up expanding all the way to using planning assumptions alongside traditional predictive planning methods. This is a time where demand planners have a tremendous opportunity to demonstrate the value of the human component in planning.
VP, Team Manager,
Gartner Supply Chain