How to maintain accurate safety stock and prevent stockouts

Calculating safety stock should be a regular exercise in any ecommerce business. Only with up-to-date numbers can you prevent stockouts.

The real challenge, however, is keeping those numbers accurate enough to actually prevent stockouts.

Often, companies calculate safety stock once. They do the math, set the stock, and move on. But as circumstances change, the original safety stock number becomes outdated. An equally impactful issue is treating safety stock as one calculation for the whole catalog. A steady seller and a seasonal item carry different risks, and a single formula leaves you over-stocked on some items and short on others.

This guide focuses on how to use safety stock calculations in real scenarios. Detailed formulas for calculating safety stock are available in the full safety stock guide. Here, we focus on applying those calculations.

Everything starts with choosing the right safety stock formula.

Safety stock calculations are quick once you know the basic inputs. But there are six main formulas to choose between, from a simple fixed days-of-stock method to more complex ones that consider changes in both demand and supply. The safety stock guide has detailed examples of all six, but let’s look at the simplest for now: 

Safety stock = Z × σLT × Davg

In this formula:

• Z is the service level factor (taken from a simple conversion table)
• σLT is the amount delivery times vary (calculated using real lead times)
• Davg is the average daily sales

For example, a 95% service level means a Z-score of 1.65. If delivery times vary by four days and average daily sales are 35 units, the calculation is 1.65 × 4 × 35, giving 231 units of safety stock for the product.

Mistakes usually come not from the math but from how the formulas are used.

Which variable is causing your stockouts?

The basic formula assumes delivery time is the main source of uncertainty. For many online businesses, this is true because customer demand stays steady while supplier delivery times change, which can cause stock shortages.

In some cases, changes in customer demand cause the unpredictability. Sales may be generally steady, but can jump seasonally, during sales, etc. In these cases, the delivery-time formula underestimates needed safety stock because it ignores sales changes. 

The best way to determine the proper formula is to look at that particular product’s sales history. Check recent delivery times and how much they vary, then do the same for daily sales over a similar time. The factor that changes more should guide your safety stock calculation. 

If both delivery time and demand vary a lot, like with seasonal or sale products, use a formula that includes both and keep a bigger buffer than simple methods suggest.

You can find the six main formulas in our comprehensive safety stock guide.

A common mistake is using one formula for all products to keep things simple. Products with steady sales and those with big seasonal changes have different risks. Using the same method for both can cause too much buffer stock for some items and not enough for others.

How high should your service level go?

People often respond to repeated stockouts by keeping more of everything. This can help at first, but the cost of extra stock adds up. 

Raising service levels from 95% to 99% can push your stock levels so high that they do more harm than good. It means higher storage costs, more space used, and more money tied up in inventory.

For most products, the service level will fall between 90% and 98%, depending on profit and how easy it is for customers to find alternatives.

Group products that can use the same formula.

It makes sense to group your products based on how predictable demand is and how much a stockout would cost you. 

For your best-selling, most profitable items, it’s worth keeping a larger buffer—a 99% service level is perfectly reasonable here. 

On the other hand, items that don’t sell often or have low margins—especially if customers can easily swap them for something else—don’t need as much protection. Keeping a huge buffer for items that rarely sell just ties up money that could be better spent on your faster-moving stock.

In any case, you should regularly revisit your safety stock calculations.

For a small handful of predictable items, a fixed safety stock is generally fine. But for almost everything else, a fixed approach is a recipe for trouble. 

Even though it’s going to be a bit more work on your end (unless you have a system to automate it), you’re much better off regularly updating your safety stock levels to keep pace with shifts in customer demand and shipping times.

What happens when you’re late to notice a stockout?

Most stockouts happen because new inventory orders weren’t placed on time.

Safety stock and reorder points must work together. Safety stock is a reserve that should stay untouched, while the reorder point tells you when to order more before you start using your reserve. If you don’t reorder on time, safety stock gets used up, raising the chance of running out due to late shipments or sudden demand increases.

The reorder point is essentially your ‘time to buy’ signal. You calculate it by taking your average daily sales, multiplying it by your delivery lead time, and adding your safety stock buffer on top. 

For example, if you sell 35 units a day with a nine-day lead time, and your safety stock is 231 units, your reorder point is 546 units—(35 x 9) + 231. With this structure in place, you can rely on data to drive your reorders.

You prevent stockouts more reliably when reordering is triggered automatically. For example, Choc on Choc, a handmade chocolate brand, uses real-time alerts to watch stock and prioritize making more product before running out. 

Manually tracking hundreds of products often leads to mistakes and missed reorder points. Automated systems like Linnworks use live sales and supplier data to flag what needs restocking before stock falls below the buffer. 

The data behind safety stock.

All safety stock formulas depend on past sales and delivery time data. If either is wrong, the buffer will be wrong too, giving a false sense of accuracy. For example, if demand forecasts are always 20% too high, the safety stock number will be exact but wrong.

Safety stock smooths the normal, random swing around a sound forecast. But that’s all it does. A forecast that anticipates the same thing every cycle stays wrong, and a buffer stacked on top of it only adds carrying cost. 

Forecast accuracy is one input. Knowing what’s actually on your shelves is the other.

In the 2026 State of Commerce Operations report, only about a third of retailers said they had excellent stock visibility across channels and warehouses, meaning two-thirds had known gaps. When you have multiple sales channels, those gaps mean you’re missing sales opportunities. 

If your system shows 80 units but the shelf has 60, your reorder point will always be late, no matter the formula. Linnworks keeps counts updated across channels and warehouses in real time, so the number you use matches what’s actually on the shelf, which helps prevent stockouts more than just improving the math.

Calculate on a schedule.

Demand, suppliers, and seasons change, so a buffer set in February might be wrong by August.

Teams that avoid stockouts check and update the number regularly. Checking every three months works well for most products, and more often for items that change a lot or are seasonal.

Doing this manually for many products isn’t practical, so if you have a diverse catalog, you’ll want software to recalculate buffers and reorder points as data changes. Linnworks stock forecasting adjusts to current sales and, combined with product-level reports that highlight changing demand and delivery times, makes updates regular instead of a last-minute rush every few months.

To assess your safety stock this week, take your top 20 products by sales, gather the last 90 days of sales and recent purchase orders for each, and recalculate safety stock and reorder points using current data. Compare these to what your system shows now. The differences are the stockouts you haven’t faced yet.

If you want a number on what they’re costing you, the inventory savings calculator estimates the lost sales and excess stock in your current setup.

FAQ

How often should I recalculate safety stock?

Quarterly is a reasonable default for most SKUs, and more often for anything volatile or seasonal. A safety stock level reflects conditions at the time you set it. Demand trends move, suppliers change, and a buffer that was right in February is often wrong by August. Teams that stay ahead of stockouts review the number on a schedule rather than treating it as a one-time setup.

What’s the difference between safety stock and a reorder point?

Safety stock is the reserve you want to avoid dipping into. The reorder point is the inventory level at which you place a new order, while the reserve is still intact. They work together: if you miss the reorder point, your safety stock is consumed during lead time, and any additional problem, such as a late shipment or a demand spike, leaves you out of stock.

Should I use the same safety stock formula for every product?

No. The standard formula assumes lead time is your main source of uncertainty, which is appropriate for products with steady demand and unpredictable suppliers. For products where demand itself swings hard, you need a formula that accounts for demand variability instead. Using a single method across the entire catalog tends to over-buffer stable items and under-buffer volatile items.

Why am I still getting stockouts even though I have safety stock?

Usually one of three things. Your reorder point isn’t firing in time, so the buffer is consumed before new stock arrives. Your inventory data is off, so the system thinks you have more than the shelf actually holds. Or your formula is protecting against the wrong variable, accounting for lead time when demand is what moves. A buffer only works if the trigger and the underlying numbers are both accurate.

What is fixed safety stock, and when does it make sense?

Fixed safety stock is a flat quantity you set and rarely revisit. It works for a small number of stable SKUs where the cost of recalculating isn’t worth it. For everything else, the level should reflect each SKU’s own demand and lead time rather than a blanket rule, because a static number drifts out of line as the business changes.

Does a higher service level always mean fewer stockouts?

Yes, but with sharp diminishing returns. Moving from a 95% to a 99% service level can nearly double the safety stock you carry, because you are protecting against rarer events. For most SKUs, the sensible range is 90% to 98%, determined by margin and product replaceability. Aiming for 100% leaves capital tied up in inventory, which you may have to mark down later.