Safety Stock Per Item: The Errors Behind Both Stockouts and Dead Inventory
Safety stock exists to absorb uncertainty — in demand, in supply, or both — so that normal variation doesn't turn into a stockout. Get it right per item and you protect service without drowning in inventory. Get it wrong, and you manage to do both at once: run out of the parts that matter while warehousing the ones that don't.
The disruption of late 2020 and early 2021 punished sloppy inventory policy hard. Lead times that used to be steady became erratic, and the comfortable blanket rules many teams relied on stopped working. The mistakes below are the ones that surfaced fastest under that strain — and they are worth fixing in calmer times too.
Common mistakes in sizing safety stock
One blanket rule for every item. "Keep two weeks of everything" feels fair and is almost always wrong. A cheap, predictable item and a critical, erratic one need completely different buffers. Blanket rules overstock the easy items and understock the hard ones.
Ignoring demand and lead-time variability. Safety stock is fundamentally about variability, not the average. Two items with the same average demand but very different swings need very different buffers. If your method only uses averages, it isn't really sizing safety stock.
Forgetting that lead time itself varies. Many calculations buffer only against demand variation and assume supply arrives on time. In 2021 that assumption broke spectacularly. When lead times wobble, that variability often drives more of the required buffer than demand does.
Applying the same service target everywhere. A 99% service level on a part that halts production may be right; the same 99% on a low-impact consumable just freezes cash. Match the target to the consequence of running out.
Setting it once and walking away. Demand patterns and supplier reliability drift. A buffer that was correct last year may be far too high or dangerously low now. Safety stock needs periodic review against fresh data.
Sizing it per item, sensibly
The goal is to spend your inventory dollars where they buy the most protection. That means differentiating items rather than treating the catalogue as one undifferentiated pile.
Segment items first — by value and by how much a stockout hurts (an ABC or criticality cut). High-impact, high-variability items earn bigger buffers.
Size the buffer from the variability of demand AND of lead time, and from the service level you actually need for that item.
Set service-level targets by consequence: protect the parts that stop the line, relax on the easily substituted ones.
Review the parameters on a schedule, especially when lead times are unstable, and adjust as suppliers recover or slip.
A useful mindset: safety stock is insurance, and like any insurance you should pay more to cover the losses that would really hurt. Spreading the same premium evenly across every item — the blanket rule — leaves your most painful exposures underinsured while you over-pay on trivial ones.
Done item by item, safety stock stops being a blunt cost and becomes a targeted defence: more protection where a shortage is expensive, less cash locked up where it isn't. In volatile conditions that distinction is the difference between resilience and a warehouse full of the wrong things.
If shortages and excess inventory somehow coexist in your operation, the fix usually starts with smarter, item-level buffering and supplier terms. XNM's procurement, sourcing & contract management can help you set inventory policy that protects service without tying up cash.