Consistency Counts - Lead-times VS Delivery On Time

A common point of contention in many businesses revolves around the balance between short lead-times from the Distribution Center (DC) to the destination point (typically a store) and the ability to meet agreed upon on-time service levels.

Non-Supply Chain functions often advocate for shorter lead-times to maximize availability and sales performance. However, Supply Chain functions are acutely aware of the costs and disruptions involved in constantly expediting and changing tasks to meet targets that often seem just out of reach.

When considering the normal distribution curve of sales volatility, the solution might seem counterintuitive. This is because, to achieve a specified stock service level at the destination point, the standard deviation of delivery on time (DOT) plays a more significant role than the lead-time itself.

In the graphic below, the x-axis represents the agreed-upon lead-time, which the inventory system uses for replenishment. The y-axis shows the standard deviation of actual lead-time performance, where a higher standard deviation indicates greater volatility. The z-axis indicates the stock required, based on the x and y axes, to meet the targeted stock service level at the destination.

As the lead-time from DC to destination increases, so does the required stock to compensate for sales demand volatility, including review and delivery frequency. Longer order-to-delivery durations increase the potential for unexpected demand and, consequently, the amount of stock needed as a buffer.

Consider a scenario where the lead-time is set at 12 days, and the standard deviation of actual lead-time is 0.7, indicating reliable lead-time adherence. The inventory required in this scenario is slightly above 20 units.

However, if this operation were to reduce its lead-times to, say, 5 days, but as a result, experience fluctuations ranging from just meeting to missing the target by up to a week, the lead-time standard deviation could spike to 2. The resulting required inventory to prevent running below the service level could jump to mid-30 units.

Lead-time and lead-time standard deviation performance and its effect on stock holding required to meet a specified service level

To provide context, imagine our operation has a 100% forecast accuracy, allowing us to perfectly predict demand over the time horizon. This scenario would effectively make the lead-time irrelevant in terms of stock holding at the destination, with lead-time variability (DOT) becoming the sole driver for additional stock requirements.

The second graphic illustrates this scenario.

Oliver.Blombery@SCIPSolutions.com.au