You reordered on schedule, the way you always do, and you still ran out of a critical item the week demand spiked. The order was placed on time. The shelf was empty anyway.
That gap is what safety stock exists to close. Demand and supplier lead times both move, and a reorder point set for an average week leaves nothing in reserve when either one turns against you. One late delivery or one busy week, and a routine item becomes a scramble.
This guide gives you the safety stock formula, a worked calculation you can copy, and a simple way to keep that buffer stocked once you've set it.
Safety stock is the extra inventory you hold beyond expected demand to cover variability in demand and lead time, so a spike in orders or a late delivery doesn't turn into a stockout. It's the cushion between "we planned for a normal week" and "this week wasn't normal."
A quick example: if you sell about 100 units a week and a bad week can run 40 units higher, holding roughly 40 units of safety stock keeps you covered when that week arrives.
It's worth separating safety stock from cycle stock. Cycle stock is what you expect to use between orders, the working inventory that turns over in the normal course of business. Safety stock sits underneath it as protection. The next sections give you the formula for sizing it and a worked calculation.
The standard safety stock formula is:
Safety stock = Z x Standard Deviation of Demand x Square Root of Lead Time
Z is the service-level factor, the standard deviation measures how much demand bounces around, and lead time is how long replenishment takes. That's the version most inventory teams start from, and it needs no special software to apply.
A few common variants are worth knowing:
Basic (fixed) method: Multiply the difference between maximum and average daily usage by lead time. It's quick and works when your demand and lead times are fairly stable.
Average-max method: Compare maximum daily usage over maximum lead time against average usage over average lead time. It's a good fit when you have usage figures but limited statistical data.
Statistical (service-level) method: The Z-based formula above, which is the most precise when you can measure demand variability.
Each formula leans on the same handful of inputs: average daily usage, lead time, how variable your demand is, and the service level you're targeting. Define those clearly and the math falls into place.
Your service level is the share of demand cycles you want to cover without a stockout, and it maps to a Z value you plug into the formula. For example:
A 99% service level is about 2.33
A 95% level is about 1.65
A 90% level is about 1.28
Higher service levels hold more cover, which protects against stockouts but ties up more cash in inventory. The right target balances how painful a stockout is against what the extra stock costs to carry.
Gather your usage and lead-time figures, pick a service level, and apply the formula. Here's the sequence, with one worked example carried through.
Pull your history: Collect daily usage and lead-time data for the item. Say the item averages 20 units a day with a demand standard deviation of five units, and a lead time of nine days.
Measure demand variability: Calculate the standard deviation of daily usage. In the example, that's five units.
Set a service level: Choose your target and read off its Z value. At 95%, Z is about 1.65.
Apply the formula and round: Safety Stock = 1.65 x 5 x Square Root of 9, which is 1.65 x 5 x 3, or about 25 units. Round to a whole stocking unit and hold 25.
Keep this calculation clean and self-contained so anyone on the team can repeat it with their own numbers.
Safety stock and the reorder point work together, but they answer different questions. Safety stock is the buffer you hold; the reorder point is the stock level that triggers a new order, and it already includes your safety stock.
Safety stock: The reserve that covers variability in demand and lead time.
Reorder point: The level at which you place the next order so replenishment arrives before you run out.
The relationship is straightforward: reorder point = (average daily usage x lead time) + safety stock. Using our previous example of an average 20 units a day with a demand standard deviation of five units and a lead time of nine days, that's (20 x 9) + 25, or 205 units. When stock drops to 205, it's time to reorder.
The right amount balances the cost of holding stock against the cost of a stockout, and it rises with demand variability, lead-time variability, and the service level you commit to. There's no universal number; the answer comes from your own data.
Segment by value and criticality: Hold more only where a stockout hurts, and keep buffers lean on low-impact items.
Recompute as demand shifts: A figure set last year drifts as sales patterns and suppliers change, so revisit it on a schedule.
Watch holding cost on slow movers: Cover that sits untouched for months is cash on a shelf, so trim it where the risk is low.
Supply chains rarely move in a straight line. McKinsey noted that most companies responded to recent volatility by increasing inventories, a reminder that buffers exist because demand and lead times shift in ways a single average can't capture. Sizing that buffer well starts with understanding your own variability, which is why solid demand forecasting matters, and why planning for supply chain disruption belongs in the same conversation.
Once you've sized the buffer, the harder part is keeping it stocked without a person remembering to reorder every cycle. A few tools can help take that off your plate.
Replenishment services: Where it's available, Amazon Business Restock can help keep critical employee supplies stocked by monitoring supply inventory levels at your work sites on designated shelving units or in vending machines and vending lockers. Amazon Business Restock is a managed service currently offered in select U.S. cities.
Your own usage data: Amazon Business Analytics can help you pull your average usage and lead-time history, the exact inputs the formula needs.
Buying the buffer: For the initial cover or a top-up, bulk buying can help you stand up stock at volume from business sellers.
These support how you purchase and replenish, but they don't calculate your safety stock for you—that number is still yours to size.
Safety stock is a calculated buffer, not a guess. Size it with the formula, tie it to a service level you can defend, and build it into your reorder point so the math does the remembering for you.
The one habit that keeps it working is recomputing it as demand and lead times change, rather than setting it once and forgetting it. If you'd rather not track every reorder by hand, where it's available Amazon Business Restock can help keep critical items stocked by monitoring inventory at your work sites, so replenishment leans on the program rather than someone's memory. The safety stock number, though, is still yours to set and revisit.
Get started today
Was this helpful?