Delivery lead time planning: Averages are not enough—variability matters too

Distinguish average duration from delivery commitments

When planning delivery lead times, the key question is not just “How long does transport usually take?” You also need to answer “How much delay can our business absorb?” The approach proposed in this article is to set a time buffer by assessing typical duration, variability in durations, and the business impact of delays separately. This keeps the buffer from becoming an undefined allowance habitually added to every order.

The arithmetic mean is the sum of recorded durations divided by the number of observations. The median indicates the central position in an ordered set of observations. If the number of observations is odd, the middle value is used; if it is even, the arithmetic mean of the two middle values is used. It is not the same measure as the arithmetic mean of all observations. The World Bank’s import lead time indicator specifically uses the median. A “typical duration” in a data table should therefore not be interpreted directly as an average or a delivery guarantee.

Variability shows how flexible the plan needs to be

The average summarizes the center of the durations; variability concerns how far observations differ from one another. For planning purposes, it is important to examine the late-delivery side of the distribution as well as its center. Therefore, even when the central duration appears similar, using the same buffer without examining the spread of delivery records may not be an appropriate starting point.

In your own shipment reviews, keep late-completed shipments visible alongside the mean and median. Rather than looking only at successfully completed, trouble-free files, track delays and cargo that has not yet been delivered separately. Instead of deleting exceptions, flag them by cause; do not treat cargo that the customer prepared late as the same problem as cargo held up during transport.

First define which duration you are measuring

Before calculating a buffer, document the start and end events. As a planning recommendation, you can use the time between cargo being ready for dispatch and receipt by the consignee as the main measure. Keep booking, pickup, terminal acceptance, departure, arrival, and final delivery records as explanatory stages within this overall duration.

The World Bank publishes the frequency with which shipments reach consignees within the scheduled or expected time as a separate logistics indicator. Tracking duration and on-time arrival separately is also fundamental to the approach this article recommends for businesses. However, historical country-level LPI indicators should not be used in place of current shipment records for a specific Türkiye–Europe route.

Create comparison groups by origin and destination, transport mode, service scope, and transfer structure. Do not treat arrival at an airport and delivery to a factory as the same end event. Ensure that the choice of calendar days or business days is consistent across all records.

Set the time buffer according to business impact

As a practical method, rank comparable past shipments by duration. First determine the delay risk the business can accept, then select a planning duration consistent with that preference. Treat the difference between the selected duration and the typical duration used as the baseline as the time buffer. This is not a recommendation for a specific number of days, but a decision framework that needs to be tested against business data.

For automotive and aerospace parts, base the assessment on the need for use before considering the part’s value: When is the material needed, is alternative stock available, and which activity could a delay stop? Develop separate decision rules rather than applying the same risk tolerance to flexible stock replenishment and a production-critical part. Having purchasing, production, and logistics teams jointly review these rules within supply chain planning is a useful way of working.

Establish an intervention rule alongside the buffer

Instead of adding more time for every uncertainty, document in advance the circumstances in which an alternative plan will be considered. Identify the responsible person and decision point for cases where cargo is not ready, the planned departure changes, or a critical delivery requirement is brought forward.

If air freight is to be considered as an alternative to an ocean freight plan, include acceptance conditions and door-to-door feasibility in the checklist, not just the main transit time. Compare the buffer with actual deliveries at regular intervals. Do not draw definitive conclusions about reliability from a small number of observations; the aim is not to promise to eliminate delays, but to make uncertainty visible and manageable.

Sources

Analysis and translations: Avemes Editorial. Source institutions are not responsible for this analysis or its translations.

• Lead time to import, median case (days) - Turkiye — World Bank; Turku School of Economics; n.d.; 2007–2018. CC BY 4.0
• Logistics performance index: Frequency with which shipments reach consignee within scheduled or expected time — World Bank; n.d.; 2007–2022. CC BY 4.0