Focus on improving forecast accuracy may lead to perpetual stockouts. Be careful!

Let me start with the companies which run their supply chains based on monthly forecasts. Most of these companies use a certain measure of Forecast Accuracy such as wMAPE, and their teams have an objective to improve it. Let’s see what happens to the behaviour of these teams.

Forecast accuracy is typically measured as a comparative metric between sales forecast and actual sales. Let’s take a simplistic case to illustrate the point. Suppose the company is selling 100 units a month consistently with customer order fill rate of 80%. The true demand is, therefore, 125 units. What should the demand planning team forecast?

The right thing to do is to forecast a demand of 125 units and work towards improving the backend supply chain to fulfil it. However, the prevailing inefficiencies may continue to hamper supplies in the short run and fill rate may continue to hover around 80%, resulting in sales of 100 units only. What does it do to the forecast accuracy as measured currently? FCA will take a severe beating and demand planner will be questioned for the drop.

As a result, demand planners play safe and take current level of sales as a surrogate for customer demand. Accordingly, they would forecast 100 units and achieve it. This action, however, will perpetuate the low fill rates.

The root cause of this system inertia, which repels any attempts to improve it, lies in the faulty measurement system causing such a behaviour. So long as the focus continues on improving forecast accuracy, rather than fill rates, the company will continue to suffer.

Is your company suffering from the same inertia?