What Is Safety Stock?
Visual input: Stocked warehouse shelves.
Safety stock refers to the reserve of products you keep on hand that acts as a buffer to prevent stockouts.
How it works is that a manager determines how many items are needed to keep fulfilling orders while inventory is replenished (either by ordering or manufacturing more products). So if something happens, safety stock ensures you always have items there to keep your business moving, making you prepared for:
- An increase in demand
- A disruption in the supply chain
- Inaccuracies in demand forecasting
- Missed or late reorders
- Damaged or unsellable merchandise
- Budget constraints that limit how quickly a business can restock
Why Safety Stock Is Important
Visual input: Empty or near-empty shelf.
As already slightly touched upon in the previous section, safety stock is important in inventory management because there’s a never-ending list of things that can affect your forecasts and supply chain.
By knowing how to calculate safety stock, you’re protecting your business from stockouts and the losses that come with them (not just financial losses, but brand-reputation damage too). But it doesn’t mean that everything in your warehouse needs safety stock, as holding too much of one type of inventory can be just as damaging to your business as holding too little.
Knowing When You Need Safety Stock
You'll need to determine which of your products you can't afford to run out of, such as your best sellers and any items necessary for keeping your business compliant.
Another factor that warrants keeping safety stock is high demand combined with fluctuating lead times, creating an unpredictable risk of a stockout at any given moment. Other benefits of learning how to calculate safety stock and implementing it include:
- Prevents stockouts and lost sales — Keep customers happy and revenue flowing by having shelves stocked, even when demand exceeds your forecast
- Buffers against supplier delays — Weather, transportation issues, customs, and, in the case of COVID-19, even global pandemics. Anything can happen with the supply chain or while your goods are in transit, that causes delays to your purchase orders
- Keeps production and fulfillment running — Nothing is worse than idle machines and half-finished builds cluttering the shop floor. Safety stock gives you enough items to prevent downtime.
Understanding When Safety Stock Is Costing You Money
Not all products are created equal, and holding safety stock for items that don’t need it (slow-moving or easy-to-source) only takes up space that could be better used for products that do sell.
Implementing safety stock for every item in your inventory is going to lead to problems such as:
- Ties up working capital — Items that are sitting on a shelf doing nothing are wasting space that could be used for growing the business
- Raises carrying costs — If everything has safety stock, you’ll need more space to hold buffer items, which means finding more warehouse space, purchasing insurance, and paying handling fees for inventory with no return on investment.
- Risks like spoilage and obsolescence — Some companies (like food and beverage, and electronics) carry items that have a certain date before which they completely lose their value, and need to be sold sooner rather than later
- Masks root-cause problems — identify problems like poor supplier performance or inaccurate demand forecasts
How to Calculate Safety Stock
Visual input: Person using a calculator and reviewing a spreadsheet at the desk.
Now that you know what it is and how not to go overboard with creating a buffer for all your inventory items, the next step is to start learning how to calculate your safety stock.
But, as the saying goes, there’s more than one way to skin a zucchini.
There are actually multiple methods for calculating safety stock, and it all depends on the type of uncertainty you’re dealing with:
- Demand
- Lead time
- Both
Most of the methods below assume you have some sales and delivery data you can use, but others also target service level and are expressed as a Z-score, which basically means the percentage of demand you want to fulfill before you run out of safety stock.
So, get your notepads and pens, because the following section is a doozy.
1. Basic Max-Minus-Average Method
The easiest method out of the bunch, it basically compares a worst-case scenario against your typical inventory turnover:
Safety stock = (Maximum daily usage × Maximum lead time) − (Average daily usage × Average lead time)
So, let’s imagine you’re a business that sells 20 units a day, but some days can see as many as 35 sales of the unit with no particular reason. And for this particular unit, the supplier’s lead time is 7 days, but there have been instances where it can take up to 10 days. Using the formula, this is how to calculate safety stock:
Safety stock = (35 × 10) − (20 × 7) = 350 − 140 = 210 units
The business would need 210 units as a buffer in case of worst-case scenarios, such as a high volume of sales and supplier lead times longer than usual. The great thing about using this method is that it’s simple to implement, ignores calculating statistical variability, and treats any historical peak as the benchmark.
2. Standard Deviation and Service-Level Methods
This method calculates safety stock for a target service level rather than letting a one-off worst-case scenario dictate everything.
And for these methods, there are several different formulas you can use. But before we delve into them, we first have to talk about the Z-score, which we mentioned earlier in the article. The Z-score shows you how many deviations (the fluctuation of unit sales each day) above average you need to cover to reach your service level (which means, as a percentage, how much stockout risk you're willing to accept). So, if you want to operate at a 95% service level, that means for 1 in 20 inventory cycles, you'll allow for one stockout.
So, to use an example, imagine a product with an average daily demand of 50 units and a standard deviation of 10 units — meaning demand typically swings up or down by about 10 units from that average on a given day. If you want to run at a 95% service level, you look up the Z-score that corresponds to it: 1.645. Multiply that by your standard deviation, and you get your safety stock:
Safety stock = 1.645 × 10 = 16.45, rounded up to 17 units
That 17-unit buffer is what you'd hold on top of your normal cycle stock to make sure you're only running out roughly 1 in 20 cycles, in line with the 95% target.
Push the target to 99%, and the Z-score jumps to 2.33 — not a small step up from 1.645, but a meaningfully larger one, because covering that extra 4% of risk means reaching further into the thinning tail of the demand curve. At the same standard deviation:
Safety stock = 2.33 × 10 = 23.3, rounded up to 24 units
So chasing a higher service level doesn't cost you proportionally — it costs you increasingly more the closer you push toward 100%, since each additional percentage point of coverage eats into an ever-smaller sliver of remaining probability.
With the core idea in place (Z-score × standard deviation gives you the buffer needed to hit a target service level), the named formulas below apply that same logic with small variations, mostly around how they treat lead time:
- Heizer and Render — Safety stock = Z × σLT, where σLT is the standard deviation of lead time demand. This is the most direct application of the Z-score concept, but it doesn't isolate time as a separate variable.
- Greasley — Safety stock = Z × √LT × σd, which factors in lead time explicitly alongside the standard deviation of demand.
- King — Safety stock = Z × √(PC/T1) × σD, where PC is the total performance cycle (including transport time), and T1 is the time increment used to calculate demand's standard deviation. King's method was built to correct for inaccuracies in demand data that simpler formulas miss.
3. Variable Demand, Fixed Lead Time
Use this approach when supplier delivery times are reliable, but customer demand fluctuates:
Safety stock = Standard deviation of demand × √(Average delay)
This method suits industries with short product life cycles, such as fashion or consumer technology, where demand can swing sharply even when suppliers are dependable.
4. Fixed Demand, Variable Lead Time
This is the inverse case — demand is steady, but delivery times aren't:
Safety stock = Z × Average sales × σLT
This formula is most relevant for businesses facing supply-side uncertainty, such as customs delays or inconsistent freight schedules, where the risk sits with the supplier rather than the customer.
5. Both Demand and Lead Time Variables
When both variables move independently, more advanced formulas combine their effects:
- King (combined) — Safety stock = Z × √[(PC/T1 × σD²) + (σLT × Davg)²], used when demand and lead time variability are independent of each other but both follow a normal distribution.
- McKinsey and Company method — σC = √[R(σS²) + S²(σR²)], where R is the average replenishment cycle, S is the average daily sales, and σR and σS are their respective standard deviations. This method is built from actual sales and replenishment-cycle data rather than assumed distributions, and calculates the stock needed to cover roughly 68% of demand scenarios (one standard deviation).
6. Complementary Calculations
Two related safety stock formulas often get used alongside safety stock rather than in place of it. Economic order quantity (EOQ = √(DS/H)) determines the ideal order size to minimize ordering and holding costs. Reorder point (Average daily demand × Lead time + Safety stock) determines the inventory level at which a new order should be placed, ensuring replenishment arrives before safety stock itself is depleted.
How to Choose the Right Safety Stock Method
Which safety stock calculation to use depends less on preference and more on what your data can support.
The right method depends on your demand stability, supplier reliability, sales history, and how many SKUs you're managing — pushing for statistical precision without the underlying data just produces a precise-looking number built on guesswork.
Match the Method to Your Situation
A few factors should drive the choice directly:
- Demand variability — Stable, predictable demand needs little more than a simple buffer. Seasonal or promotional spikes call for a method that explicitly captures the gap between average and peak demand.
- Lead time reliability — Short, consistent lead times reduce the need for a large buffer. Long or unpredictable lead times (especially with overseas suppliers or customs exposure) require a method that accounts for lead-time variation directly, not just demand variation.
- Service level targets — Higher service commitments, like next-day fulfillment promises, require a Z-score-based method, since only the statistical approaches let you target a specific percentage of demand covered.
- Reorder frequency — Ordering frequently allows for a smaller safety stock, since you're never far from your next replenishment. Infrequent ordering requires more buffer to cover the longer stretch between orders.
- Cost tolerance — Every method eventually runs into the same tradeoff: higher service levels and heavier buffers cost more in tied-up capital and storage. The method you choose should reflect how much of that cost the item's importance actually justifies.
A Decision Framework by SKU Count and Data Maturity
Beyond these factors, the practical constraint is usually data infrastructure rather than math. The table below matches typical business situations to the appropriate method:
The dividing line across this table is data, not mathematical complexity. The statistical and combined-variability methods require clean, per-SKU historical demand and lead-time records across every channel a business sells on — something spreadsheets struggle to maintain accurately much past 50 SKUs or multiple storefronts. The safety stock formulas themselves don't get harder to apply as the record-keeping behind them does.
A Data Readiness Checklist
Before upgrading to a more advanced method, it's worth confirming the underlying data can actually support it:
- Is there at least 8 to 12 weeks of clean daily demand history per SKU?
- Are lead times recorded per supplier, reflecting actual variation rather than a single estimated average?
- Is demand data unified across all sales channels, or fragmented by storefront?
- Does the recorded on-hand count match the physical shelf, or has it drifted from manual corrections?
A "no" to any of these is a signal to fix the underlying data before adopting a more advanced safety stock formula.
A clean, simple calculation built on reliable data will outperform a statistically sophisticated one built on poor data every time.
Safety Stock vs. Reorder Point: How They Work Together
Safety stock and reorder point work closely together, but answer different questions.
Safety stock is the quantity of extra reserve inventory held to absorb unexpected demand or supply disruptions. Reorder point is a threshold — the inventory level that signals it's time to place a new order. Safety stock is one of two components in the reorder point calculation, alongside expected demand during lead time.
Cycle stock, the demand expected during a normal lead time, gets consumed first. The reorder point marks the moment to reorder before it runs out, and safety stock is what remains if the new order is delayed or demand exceeds expectations.
Neither safety stock nor reorder point replaces the other.
Safety stock protects against running out of product, while the reorder point protects against timing errors, as ordering too early ties up cash and ordering too late risks a stockout. The most common breakdown in this relationship is inaccurate demand forecasting, which miscalibrates both numbers simultaneously, often compounded by supply chain delays that silently push lead times beyond what the reorder point assumes. Inventory management software that flags forecasting drift or delivery delays early helps catch this before a stockout occurs.
Using ERP Solutions to Manage Safety Stock
ERP systems replace static, manually maintained spreadsheets with a live, centralized view of inventory.
Instead of planners recalculating buffers by hand, the system tracks stock movements as they happen, applies demand forecasting to anticipate future needs, and adjusts safety stock and reorder thresholds automatically as demand or supplier performance shifts. This shifts inventory management from a reactive, intuition-driven process to one grounded in real-time data — reducing both stockouts and the excess carrying costs that come from overcorrecting for them.
Digit takes a simpler approach.
It doesn't have a dedicated safety stock field. Instead, the same function is handled through the Minimum Stock Level setting on an individual item, set within that item's Inventory settings. Once a threshold is entered, Digit sends a daily email alert whenever stock drops below it, giving a warning before the item actually runs out. It's a manually configured version of the automated buffer described above, relying on the number remaining accurate over time rather than recalculating dynamically.
Want to see for yourself? Book a call with one of our inventory management experts who can take you through Digit and answer any remaining questions you may have around how to calculate safety stock.
