In this blog, we will consider the second, more challenging source of safety-stock-related variation: Lead-time.
Safety stock addresses variation with unpredictable timing. In a process (such as replenishment lead time) that is reasonably stable, random variation has quantifiable magnitude, and this enables using objective statistical techniques to establish and optimize safety stock.
Why do we say that quantifying lead-time variation is a challenge?
First, it may not be – if your business system records, for each receipt of each inventory item, when the replenishment signal was triggered (kanban, PO, release, work order, etc.) and when it was fulfilled – you have your source of data.
However, this data is often unavailable or unreliable.
But you can still have a good data source.
Often, the person responsible for planning or reordering an inventory item has a good experiential or intuitive sense of the item’s range of lead times. For instance, a buyer may reliably estimate that purchased item ABC’s average lead time is 20 days, its minimum lead time is at least 17 days 95% of the time, and its maximum lead time is 27 days or less 95% of the time.
Estimates such as these can provide the realistic, representative range of lead-time variation that you need for a statistics-based safety-stock analysis.
We now have the sources of data that we may use to quantify the magnitude of random (unpredictable timing) variation for safety stock.
To determine correct safety-stock levels, we must now address two key issues:
- Isolating, quantifying and statistically defining the random-variation component of demand and lead time data
- Quantifying and defining the other factors that have a significant effect.
We will cover these issues in our next blog posts….