We are going to take a more detailed look at the primary factors that affect safety stock over the course of the next 7 blogs .
Six primary factors that affect safety stock are:
• Demand Variation
• Lead-time Variation
• Replenishment Interval
• Customer Service Level
• Probability of Order Cancellation vs. Past-Due Backlog
• Cost of Expediting
Random Demand Variation – What Does Yours Look Like?
By definition, the timing of random demand variation cannot be ascertained. However, its magnitude can be objectively quantified. How?
Historical demand data typically includes trend, seasonality and special causes (identifiable events such as promotions, catastrophes, true “one-time” spikes and transaction errors). Once the historical data has been adjusted appropriately for these things, the remaining data may be represented by a distribution. It is important to consider demand variation at its most granular level, which is usually daily, rather than weekly or monthly. A typical distribution of daily demand values is shown in the figure to the right.
Importantly, and as the green curve superimposed on the graph in the figure illustrates, the distribution of demand data is nearly always right-skewed, and in many cases even more extremely than in this example. This, of course, is because the lowest demand value is never less than zero, but the highest demand value has no such absolute bound.
Demand rarely resembles the classic, symmetrical, normal distribution or “bell curve,” represented by the red curve superimposed in the figure.
Also, as mentioned earlier, demand typically occurs – and is fulfilled – daily, not weekly or monthly. As a result, any measure of actual service-level performance should reflect daily behavior