What is safety stock and how to calculate it?


1/3 of businesses will miss a shipment deadline because they’ve sold an item that wasn’t actually in stock!

What is safety stock?

Safety stock is an integral part of inventory management. Retaining adequate safety stock levels can help avoid unnecessary stock outages in a fast-paced e-commerce business setting, while calculating sensible safety stock levels can avoid holding too much stock resulting in possible cash flow issues and unnecessary costs. 
Safety stock levels can be incorporated into other inventory management tools and formulae such as the reorder point formula to create a meaningful and effective inventory management system critical to growing e-commerce and e-commerce fulfilment businesses. 

Why keep Safety Stock?

But why have safety stock – why not just keep your stock levels as low as possible? Well, safety stock defends your business against two uncertainties: demand and lead time. Standard inventory management processes, such as forecasting purchases based on reorder levels for each product, can maintain stock during ‘normal’ times; but there are times when demand increases unexpectedly, or usual lead times are not met. This is where safety stock can mitigate any unforeseen changes to the norm. 

COVID-19 has created massive uncertainty in the supply chain across a whole host of sectors and we are seeing unprecedented demands on the stock management function with lengthy, inconsistent lead times and a reduction in variety of products available from suppliers. Now more than ever there is a need to ensure that safety stock levels are in place to try and limit the impacts of these disruptions. 

How Does it Work?

Firstly, safety stock calculations are not intended to eliminate all stock outs—just the majority of them. A 95% service level will eliminate all but 5% of stock outs. A higher service level demands a higher safety stock and while it would be lovely for the inventory manager to hold enough safety stock to cover 99% of all situations it is rarely economically viable to do so. 
Safety stock calculations allow managers to choose their level of risk and exposure to poor services levels and weigh them up against inventory costs. 

The 6 Safety Stock Formulae

There are six recognised safety stock formulae. Here we aim to outline what these are, how to calculate them and when to use each. Different business types operating in different sectors have different demands on their inventory and cash flow. Getting the right safety stock formula can be the basis for an effective inventory management system. Similarly, within one company there will be many stock items with differing characteristics and safety stock levels will be affected by these, including:

  • Consistency of sales volumes – some will sell consistently over the year others will be more ‘seasonal’
  • Value – a business may sell some very high value items and some very low value – the overall value of the safety stock needs to be considered.
  • Importance of the stock item to the business – some businesses may have a stock item that is their ‘flagship’ product which must be available at all times.

Spending time on working out the right safety stock formula for your business, along with safety stock levels for different stock items will pay dividends in effective management of stock into the future.

Basic Formula

The basic safety stock formula simply takes the average sales per day along with the number of stock days you feel you need to keep stock of.

For example, if you sell 100 products a day and you feel that 5 days’ worth of stock is a good safety net the calculation would simply be:

100 (products) x 5 (days’ worth of stock)

The number of days’ worth of stock will largely depend on the lead time of the product along with the time you feel comfortable with. So, if lead times are 5 days but you would like to keep a few days extra stock just in case then the second figure would be more like 7 – this is a very simply way of calculating the safety stock and is relevant for those that do not have variable sales and for products where the safety stock levels are not that critical.
While easy to execute, such techniques generally result in poor performance. A sound, mathematical approach to safety stock will not only justify the required inventory levels to business leaders, but also balance the conflicting goals of maximizing customer service and minimizing cash tied up in inventory. 

Average – Max

If, for example, a product usually sells around 100 units per day but can occasionally be as many as 200, then an alternate formula is required. It’s important to have enough stock to cover busy sales days without holding too much stock. If the lead time has also varied in the past then this must be taken into account, ordering too far in advance can leave a lot of cash tied up in inventory but inconsistent lead times from suppliers can lead to stock outs if they aren’t handled properly. You might not need to switch supplier if you bake this inconsistency into your safety stock. For example, let’s say that the item has a lead time of 10 days from the supplier but on occasion it has been up to 15 days. 

As such we can calculate the safety stock levels for this product using the following formula:

Safety Stock =(maximum daily sales x maximum lead time in days) – (average sales x average lead time)

In our example: Safety stock = (200 * 15) – (100 * 10) = 3000 – 1000 = 2000

This is a useful formula for balancing out sales and lead times that are not uniform, but it’s important to remember some key principles when handling your data. One example would be to remove ‘anomalous’ results when looking to find your average. For example, if you were selling 100 units of Product X every day throughout 2022 except one day where you sold 12,000 – it would be reasonable to remove the day where you sold 12,000 from your calculation of the daily average sales to prevent a one-off occurrence from skewing your data.

Normal Distribution

The normal distribution method is applicable when taking into account service level only.

This model makes use of service level and normal distribution of data and can be used when the only variable is the service level rather than demand or lead time.

Essentially when you set your service level it will determine a safety factor which you can use as a safety stock which is uniform according to demand. As the chart demonstrates your probability of selling more or less is uniform. 
If the only variability you need to protect against is demand variability, and strong data is available the safety stock needed to give a certain level of protection is simply the standard deviation of demand variability multiplied by the Z-score.
So, as an example if you set a service level of 50% it means that you have a 50/50 chance of selling more or less than the ‘average’ month of sales. 

However, if you are wanting a service level of 90% the calculation will give you a multiplier figure to set your safety stock. 
A company’s ideal safety stock level will be based on its tolerance for stockouts. For example, if company leaders decide that the company can tolerate stockouts during no more than 10% of the cycles, that sets service level goal at 90%. While designing for a higher service level—say, 98% would result in fewer stockouts, this requires significantly more safety stock.
Good inventory management practice demands that Z should be set independently for stock items based on strategic importance, profit margin and value as discussed above. 

The Z score is calculated via the desired service level discussed above. This can most easily be calculated in Excel using the NORM.S.INV function as below:

Normal Distribution

Service Rate

Z =Coeff service
99.90% 3.09
99% 2.33
98% 2.05
97% 1.88
96% 1.75
95% 1.64
94% 1.55
93% 1.48
92% 1.41
91% 1.34
90% 1.28
89% 1.23
88% 1.17
87% 1.13
86% 1.08
85% 1.04
84% 0.99
83% 0.95
82% 0.92
81% 0.88

Normal Distribution with Uncertain Demand

This should be used where you are in a position where the only variability you need to protect against is demand variability, and you have strong historical data are available.
The safety stock needed to give a certain level of protection simply is the standard deviation of demand variability multiplied by the Z-score.
So, if no safety stock is required the Z score will be zero, by the calculations if the Z score is 1 then safety will cover demands for 84% of the time. The relationship is nonlinear and the higher the % required results in a disproportionate amount of Z score, and thus safety stock, required. As such some stock items may require up to 98% levels but with other items it may be that closer to 50% would be acceptable. Using this formula across each stock item will enable inventory managers ensure that key stock items very rarely go out of stock but less important ones can be kept at less secure safety stock levels. 

The formula is:
Z x Demand Standard Deviation x sqrt (AverageLT)

Normal Distribution with Uncertain Demand & Independent Lead Time

In the previous example safety stock is used to mitigate against demand variability. However, when variability in lead time becomes the primary concern, the safety stock equation changes: 
σLT=standard deviation of lead time Davg= average demand.
There are cases where both demand and lead time variability occur in the same stock product independent of each other. Calculations can be combined to give a lower total safety stock than the sum of the two individual calculations 

The formula we use here is:
Z*Average Sale*Lead Time Standard Deviation

Which formula should I use?

With such an array of formulae to choose from it can be hard to decide which method to use. The quality of data, along with the characteristics of the stock movement will largely define which to use. 

Regardless of the characteristics, if volumes are generally low it is probably sufficient, and most effective to use the average-max formula – it is relatively easy to calculate and requires minimal data to ascertain safety stock levels. 

Where sales are higher per stock item – maybe over 200 sales per month or more then assuming the lead times are relatively stable then ‘Normal Distribution with Uncertain Demand’ will be most effective. This requires less data than more complex formulae and will mitigate for most stock shortages. 

Finally, should you have issues with variable lead times then the formula “Normal Distribution with Uncertain Demand & Independent Lead Time” should be implemented. Perhaps more complicated but certainly better for those stock items that have complicated demand and lead times issues surrounding them. 

Whether you use the same formula across all stock items or pick and choose formulae across the range is an interesting question. Of course, the more complicated formulae require more complex calculation and data collation, but if the data sets are strong and uniform across all stock items then using standard calculations across all stock items can be automated. This negates any complicated division of stock into differing groups dependent on their characteristics. Across a large range of stock that may be held by e-commerce businesses this can become cumbersome and counterproductive. 

Key to a successful inventory management system is to put in place a system that works according to the data available and the needs of the business. There is little point in complicating stock management if this is quite straightforward or the business is new without sufficient data available. Whereas implementing more detailed safety stock calculations can have a huge impact across a very large range of stock where a large number of stock outages can cause significant disruption.

How to Calculate Reorder Points?

While we have covered reorder points in detail in other articles it is worth revisiting the reorder point formula here when used with safety stock. The reorder point calculation can be done with, or without safety stock levels incorporated. Whether to use safety stock levels or not really does depend on a range of factors including how critical the stock item is to the business, and how the business as a whole chooses to operate their stock. For example. Just in Time inventory management would largely operate without a safety stock figure. 

  1. When used with safety stock the reorder point formula looks like this: Average daily usage rate x Lead time
  2. (Average daily usage rate x Lead time) + Safety stock


  • Average daily usage rates = sales of that specific item each day
  • Lead time = the length of time it takes to restock the item
  • Safety stock – the amount of stock you as a business feel confident at holding for unforeseen circumstances – i.e. a sudden rush of sales, delay in the lead time, stock shortages at your supplier and so on.

So, it is relatively simple to see how this reorder point calculation can be bought into action once your safety stock levels are ascertained.

Calculating Demand 

Demand is best calculated over a fairly lengthy period of time, although not too long as stock items popularity may change over time. A manual stock calculation can be done over say 3 months:







Average per month (total divided by 3)

So, the average daily usage rate is 3000 per month, so around 100 per day.
Increasingly, e-commerce companies with differing inventory levels are using stock management software, such as Veeqo to manage their stock levels and such software can very easily be used to calculate demand levels across all your stock items.

Calculating Lead Time

Again, it is best practice to take an average over the same time period. 


Lead time





Average per month
(total divided by 3)

So, the average lead time for this stock item is 10 days.

Calculating Service Level

Much of inventory management is approached from a cost of stock point of view and is designed to keep stock at the minimum level required to satisfy the customer. Service level turns this on its head and approaches the stock level requirements from a customer satisfaction point of view. Increasingly inventory management technology is being utilised to reduce inventories while simultaneously increasing customer service levels.

In inventory management, service level is the expected probability of not hitting a stock-out during the next replenishment cycle or the probability of not losing sales. This is where operational costs meet opportunity costs. Mostly customer sensitivity to stock-outs vary from one stock item to another making the optimum service level specific to each product individually. The item that is key to your reputation – your key selling item or signature product is critical to your company’s profitability. As such service levels for this item may well be significantly higher than a product that sells infrequently and is low value to your business. 

A simple way of looking at this is to rank products in order of profitability to your business where:

  • items A, top 20% products, are classified as “critical”: high service level, e.g. 96-98%;
  • items B, next 20-30% products classified as “interclass”: medium service level, e.g. 91-95%;
  • items C, last 50-60% products classified as “trivial”: lower service level, e.g. 85-90%

However, there are many different options that can be considered if taking a more – item by item approach. Formulae can be complicated and may be beyond all but the most statistically minded and this is where inventory management software such as Veeqo can be utilised to manage and calculate service levels. 

Using Safety Stock with Economic Order Quantity

Again, we have looked more in depth at EOQ elsewhere, but it is worth revisiting the relationship between EOQ and safety stock.

The ultimate goal of the EOQ formula is to calculate the optimal number of product units to order at any given time. It is used to determine a company’s inventory reorder point such that when inventory falls to a calculated level, the system triggers the need to place an order for more units. 

Using Economic Order Quantity alongside safety stock levels can ensure that the stock levels are kept between levels that both ensure stock outages are minimised but levels are economically viable. 

Check Frequently and Regularly

We have discussed why using accurate and extensive data is important when calculating all inventory levels, including safety stock but equally important is ensuring that the levels are still relevant over time. All of the variables are just that – variable, and over time things change. Products become less popular, pricing changes, lead times can be volatile. 

Particularly now, when Covid and Brexit are heavily impacting prices, led times and availability of products making sure your safety stock levels are adequate is vital. Forecasts and current EOQ, ROP and safety stock levels need to be continually monitored so as to keep your business profitable and relevant. Fast paced e-commerce businesses rely on having strict inventory control models in place and across a wide range of inventory this can become a challenge. Using inventory management software such a Veeqo can help. Keeping all data in one place, carefully controlled with in built monitoring tools can take the pressure off inventory managers and reduce workload. The software can flag discrepancies and problems before they arise and inventory management can move from a reactionary process to a proactive and seamless operation. 

As we have discussed there are many variables to consider and within e-commerce particularly during volatile periods keeping a handle on lead times can be key to the success of inventory management controls. The reduction and stability of lead times can be a crucial factor in a businesses ability to maintain enough stock to meet demand. The lower and more reliable the lead time the les safety stock needs to be held resulting in reduced overheads in inventory value and warehouse costings. While inventory management tools are vital working with suppliers to ensure favourable lead times is key to keeping costs to a minimum. 

How you can misuse safety stock

The one big mistake companies often make when they use the standard safety stock formula is that it can overstate the safety stock required. 

In real work action the inventory figures when using the safety stock z-factor formula can be incredibly high resulting in inventory managers back-pedalling away from using the system.

This can result in inventory managers largely ignoring the safety stock figure as it is far too expensive to keep that amount of stock. But crucially they often then fail to replace the calculation with anything – largely relying on ‘gut feeling’ to set the level. This can, and does lead to significant stock outages down the line, massive variances in inventory control across the whole business and ultimately a confused and ineffective inventory management system. 

Safety Stock Decline

Inventory is a highly visible asset and, in many companies, also the largest asset. In today’s highly competitive global economy, inventory has become the focus of improvement for many companies. Cutting safety stock levels is one obvious area where savings can be made and e-commerce businesses with large stock ranges are often tempted to reduce safety stock across the board. 

Letting safety stock go to zero is clearly a highly effective way of cost reduction. However, as safety stock nears zero service levels start to drop too. For most companies, particularly in e-commerce this reduction in service levels can end up being far more costly than keeping safety stock. Lost customers, complaints, increased transport cost and more can result in businesses witnessing massively reduced profits, reputation and even failing.


There is no denying that safety stock calculations can be hard to understand and equally hard to put into practice. Inventory demands, service levels and cost pressures change and safety stock levels have to be reactive to these demands – they can’t operate within a business alone. It is critical that the inventory manager has a handle on safety stock as part of a holistic approach to stock control and sees safety stock as a tool with which to manage stock levels and costs. Using inventory software can effectively take the guess work and confusion out of inventory control and allow a more controlled and fluid stock adjustment process.


Source link