Safety Stock, Item by Item: The Difference Between Buffer and Bloat
Safety stock is the inventory you hold to absorb the variability that forecasts cannot. The hard part is not deciding to hold it — it is deciding how much, for which item. In a year of erratic lead times and price swings, a sloppy answer is expensive in both directions: too much capital frozen in slow-movers, too little of the part that actually stops the line.
The recurring mistake is treating safety stock as one policy applied to everything. A good approach sizes it item by item, against the two things that actually drive it: how uncertain demand is, and how uncertain and long the lead time is. Here is what that looks like in practice versus what goes wrong.
What good looks like
A sound safety-stock calculation starts from real variability, not a gut feeling. Good teams set a service-level target that reflects the cost of a stockout for that specific item, then size the buffer to cover demand and lead-time variability over the replenishment period. Critically, they recognize that lead-time variability usually dominates — a part with a steady forecast but a wildly unpredictable supplier needs more buffer than its smooth demand suggests.
Differentiate by item: a critical, long-lead component is not managed like a cheap, plentiful one
Base the buffer on measured demand variation and lead-time variation, not last year's number
Set service levels by consequence — protect the parts that halt production, accept more risk on the rest
Re-tune as conditions change; 2022's lead times are not 2019's
Good teams also separate the two reasons inventory runs out. A stockout from demand spiking is a different problem than a stockout from a supplier slipping three weeks, and the buffer should respond to whichever is larger for that item.
What bad looks like
Bad practice is recognizable from across the warehouse. It usually appears as one of a few habits.
The flat percentage. Holding the same fixed weeks of supply on every SKU ignores that items differ wildly in risk and cost — you overstock the safe ones and underprotect the dangerous ones.
Ignoring lead-time variability. Sizing the buffer on demand swings alone, while pretending the supplier always delivers on the promised date, is the single most common cause of being caught short in a volatile market.
Set and forget. A buffer calculated three years ago, never revisited, quietly becomes either dead capital or a guaranteed shortage as lead times drift.
Padding by fear. Adding stock because the last stockout hurt, with no link to measured variability, inflates inventory across the board and hides the real problem.
There is a quieter cost to getting this wrong in 2022. Inflated safety stock ties up cash exactly when financing is more expensive and prices are rising, while thin buffers on the wrong items leave you exposed to the lead-time shocks that have become routine. Both failures come from the same root: not sizing the buffer to the item's actual risk.
Where to start
You do not need a perfect model to improve. Start by classifying items by value and criticality, measure the demand and lead-time variability on the handful that matter most, and set their buffers deliberately. Then review on a cadence rather than once. The goal is not the most inventory or the least — it is the right inventory on the right items.
If you want a clear-eyed review of how your buffers are set and what they are costing you, XNM's procurement, sourcing & contract management can help you right-size inventory item by item.