5 Essential Sawtooth Curve Scenarios
There’s no point to create a sawtooth curve unless we know how to analyze it. In order to effectively analyze sawtooth curves, it’s imperative to familiarize ourselves with a few essential scenarios.
In that effort, below are some example sawtooth curves to help better understand how curves reflect dynamic changes.
Each of the following examples assume an item with a 10-day lead-time. Other variables are changed to illustrate their impact on the replenishment plan and the corresponding sawtooth curve.
1 Order per LT, No SS, Stable DD, Predictable Receipts
Let’s start by charting an ideal situation with no variation. The item has a consistent (stable) daily demand of 10 units. As illustrated in Figure 5, because we have perfect consistency, we have no need for safety stock.
1 Order per LT, No SS, Stable DD, Delayed Receipt
If lead-time is off by one day, so that you didn’t receive the replenishment on Day 11, you’ve got a delayed receipt situation and zero inventory OHB. Figure 6 illustrates this. The red data point below the zero on the sawtooth chart represents negative inventory.
In this scenario, not only are you missing 10 pieces on Day 11 but you’re also consuming 10 pieces on Day 12. You received 100 pieces on Day 12, but because you must account for the pieces that were consumed on Day 11, you have 80 left instead of 90. You would have had 90 if you received the replenishment on time.
There was no safety stock to cover the delayed receipt until it came in. This puts the sawtooth curve at risk and has a possible negative impact on production.
1 Order per LT, No SS, Variable DD, Delayed Receipt
In the real world, you won’t have a simple scenario where the daily demand is 10 pieces every day for 10 days. Therefore, you’ll need to consider demand variation as well as delays from the supplier.
On an inventory sawtooth curve, when the OHB varies from day to day, you’ll see wave-like patterns between the expected replenishment days.
The OHB shows red data points for days where consumption outweighed the expected daily demand. On consecutive days of negativity, the balance will continue to accumulate when there’s no safety stock.
It makes sense to have some safety stock to account for supplier delays and fluctuations in consumption.
1 Order per LT, SS = 20, Variable DD, Delayed Receipt, 120 Order Quantity, Order Every 10 Days
To address the daily demand, you might increase the 100-replenishment order to 120. The extra 20 outside of the 100 is the safety stock. Nothing else changes in the daily demand of 10. You have the 20 in safety stock to buffer supplier delays or higher consumption days.
Unfortunately, with a 10-day order schedule, safety stock always adds to the OHB whether it’s needed or not! Unfortunately, far too many inventory managers respond to situations such as this by trying to reduce order quantities, which by extension means they are reducing safety stock.
In theory, over time, this will reduce and eventually eliminate overstock. In reality, reducing order quantities may or may not reduce overstock. Of course, this all depends on future demand variability which inventory managers have no way of controlling!
What happens if demand begins to wane? Eventually we could find ourselves in another stockout situation and as a result, adjusting our order quantities back-up. This type of knee-jerk “over-steering” magnifies the bull-whip effect, driving up internal and external inventory management costs.
Luckily, there’s a better way to handle this situation. To properly counter over-ordering, you are far better off to adjust order frequency rather than order size.
1 Order per LT, SS = 20, Variable DD, Delayed Receipt, 120 Order Quantity, Consumption based Replenishment
One way to adjust order frequency would be to simply ask the supplier to deliver on less frequent schedule. However, this is akin to reducing order size (and potentially far worse).
What do we do?
We know we can’t control consumption, but we can control how we respond to consumption. The savvy inventory manager will recognize that what we really want and can reasonably achieve is replenishment in response to consumption; consumption-based replenishment.
Consumption-based replenishment is kanban.
Triggering a kanban order typically occurs in one of two ways:
- When and order quantity is entirely consumed (“empty-a-bin”)
- When an order quantity is opened for the first time (“break-a-bin”)
Figure 9 assumes that new replenishment orders are triggered when an order quantity is opened, and the first unit is consumed. Let’s walk through this example.
- Day 2: A new replenishment order is triggered on day 2, when the first item is consumed from a previously un-opened 120 piece container. (The chart in Figure 9 shows day 1 with 130 on-hand, which would equate to an un-opened 120-piece order plus 10 remaining units from a prior order.)
- Day 12: The order triggered on day 2 is due, because our lead-time period is 10 days (2 + 10 = 12)
- Day 13: Our 120-piece order arrives due to a 1-day supplier day, as in the previous examples. Because we have safety stock, the delay does not cause us to suffer a stockout.
- Day 41: We only receive 4 orders in 41 days rather than 5 as was the case in Figure 8. This is because our replenishment orders and corresponding inventory receipts are paced by consumption, rather than a static schedule.
Notice in Figure 9, that our 20 units of safety stock still prevents a stockout, and our 10-day lead-time has not changed, but because our replenishment is now triggered by actual consumption, we don’t build-up excess inventory overtime.
The takeaway from these scenarios is that inventory management requires both a strategic approach and analytical rigor. Sawtooth curves are visual tools for inventory management, well-suited to assisting in these requirements. Because they’re visual, it’s good to recall that they are not intended to be 100% accurate and are not meant to be our only tool for inventory analysis. In addition, several mathematical formulas complement sawtooth curves to help us to better analyze inventory on-hand balance. In part 4 of this 4-part series we take a closer look at a few key inventory formulas.