How Much Safety Stock Should You Hold? (Formula and Service Levels)

Hold enough safety stock to cover the gap between your worst-case demand and your worst-case supply over one lead time, no more. A common quick formula is safety stock = (maximum daily sales times maximum lead time in days) minus (average daily sales times average lead time). The statistical version sizes it to a target service level: safety stock = Z times the standard deviation of demand over lead time, where Z is 1.65 for 95% and 2.33 for 99%. The right amount is never a flat number of weeks; it scales with how variable that SKU's demand and your supplier's lead time actually are.

Too little safety stock and you stock out during a demand spike or a late shipment. Too much and you park cash on a shelf and pay to store it. Below is how to size it, the two formulas, how service level changes the answer, and why the number is different for every SKU.

What is safety stock and why hold it?

Safety stock is the buffer you keep on top of your expected demand to absorb two kinds of surprise: demand that runs hotter than forecast, and supply that arrives later than promised. If demand and lead time were perfectly predictable you would need none. They never are, so safety stock is what stands between a normal bad week and a stockout. It is the input that turns a naive reorder trigger into a reliable one, which is why it sits at the heart of the reorder point formula.

How much safety stock should I hold?

Hold the amount that covers your realistic worst case over a single replenishment lead time, sized to the service level you are willing to pay for. There are two standard ways to get the number, one quick and one statistical, and they suit different levels of data.

The quick max-minus-average method

Safety stock = (maximum daily sales times maximum lead time) minus (average daily sales times average lead time). Say a SKU sells 20 units a day on average and 35 on its busiest days, and your supplier's lead time runs 10 days normally but 16 at worst. Safety stock = (35 times 16) minus (20 times 10) = 560 minus 200 = 360 units. It is easy to compute from your own history and needs no statistics, though it can overstate the buffer because it assumes the worst demand and worst delay land together.

The statistical service-level method

Safety stock = Z times the standard deviation of demand over the lead time. Z is the service factor for your target service level: 1.65 for 95%, 2.05 for 98%, 2.33 for 99%. If demand over the lead time has a standard deviation of 150 units and you want a 95% service level, safety stock = 1.65 times 150 = about 248 units. This version ties the buffer directly to how much protection you are buying, which is the honest way to think about it, and it needs a clean read on demand variability to be trustworthy.

How do I calculate safety stock when lead time varies?

When both demand and lead time swing, use the combined formula: safety stock = Z × √(L × σ_d² + d̄² × σ_L²), where L is the average lead time in days, σ_d the standard deviation of daily demand, d̄ average daily demand and σ_L the standard deviation of the lead time. Late suppliers, not spiky demand, are usually what inflates the buffer.

A worked example makes the point. A coffee roaster's 250 g espresso bags sell 10 a day with a daily standard deviation of 4, the supplier quotes 10 days, and the owner wants 95%. With a fixed lead time the buffer is 1.65 × 4 × √10 ≈ 21 units, so the reorder point is 100 + 21 = 121. Now suppose the "10 days" really ranges from 8 to 16, a lead-time standard deviation of 2 days: √(10 × 16 + 100 × 4) = √560 ≈ 23.7, and 1.65 × 23.7 ≈ 39 units. Nearly double, from the supplier alone.

If you do not have clean daily standard deviations, a defensible screening shortcut is to hold half a lead time of demand as safety stock, then tighten toward the formula as your data improves. It scales with the two inputs that matter, demand rate and lead time, and it can be computed for a thousand SKUs with no statistics.

How does service level change how much I hold?

Every step up in service level costs disproportionately more stock. Going from 95% to 99% roughly means moving Z from 1.65 to 2.33, about 40% more safety stock, to close a four-point gap in reliability. That is why a flat 99% across the whole catalog is usually a mistake: you pay a steep buffer on slow, low-margin lines that barely matter. Set a high service level on the SKUs where a stockout loses a valuable sale or a customer, and accept a lower one on the long tail.

Why safety stock is different for every SKU

A stable SKU that sells 10 a day like clockwork from a reliable local supplier needs almost no buffer. A volatile seasonal SKU from an overseas supplier with a swinging lead time needs a large one. Both inputs, demand variability and lead-time variability, are specific to the product, so a single company-wide rule such as two weeks of cover is wrong in both directions at once. It over-protects the steady lines and leaves the volatile ones exposed. This is the case for calculating safety stock per SKU, and for demand forecasting software that recomputes it as demand and lead times drift.

Lead-time accuracy is half the calculation, and it depends on knowing what your suppliers actually do, not what they promise. Pulling real dates and quantities off your purchase records, rather than trusting the quoted terms, is what makes the buffer honest. Teams that turn supplier invoices into clean, sortable data can measure true average and maximum lead times per supplier instead of guessing, which tightens every safety stock number downstream. Once the buffer is right, inventory replenishment software can recompute it and the reorder point every week as demand and lead times shift.

Can I hold too much safety stock?

Yes, and it is the more common mistake. Excess safety stock is cash frozen on a shelf, plus the carrying cost of storing, insuring and financing it, plus the risk it ages into dead stock. A buffer sized to a worst-case that never happens looks like prudence and behaves like waste. Review your safety stock levels as demand and lead times change; a number set two years ago on last-generation lead times is almost always wrong now.

Turn the buffer into a live reorder point

Safety stock is not something you hold on its own; it is added to expected demand over the lead time to set the reorder point that triggers your next order. As demand and lead times shift, both the forecast and the buffer should move, and the reorder point with them. Doing that by hand across a full catalog is the work a forecasting layer takes over, projecting demand per SKU and keeping the buffer and the trigger current. That is what Storekeeper does on top of the counting your system already handles.

To see which SKUs are running thin on cover right now, paste your current stock and recent sales into the live stock scan at the top of the site. It flags the lines closest to a stockout, so you know where a bigger buffer earns its keep and where you are simply holding too much.