Jul 032009

The purpose of safety stock is to protect against fluctuations in demand or supply or any uncertainty.  Safety stock is based and dependent on:


  • Frequency of ordering
  • Variability of demand (compared to the forecast)
  • Desired service level


We’ll address these one at a time:


Frequency of ordering – How often you order or manufacture your products will affect your safety stock.  The more frequent your ordering the less your demand during lead time therefore the amount of replenishment is smaller than less frequent ordering.  However, due to the smaller quantities your risk of a shortage increases but exists for a shorter period of time.  Less frequent ordering decreases the risk of a shortage but stock outs last longer.  It’s been my experience that it’s better to have more frequent stock outs for shorter periods of time than to have less frequent stock outs for longer periods of time.  Customers may understand or have safety stock of their own that will be lower if any stock outs are for only a short period of time


Variability of demand – No matter how accurate your forecasting is you will experience variability in your demand.  Safety stock is primarily to absorb this variability.  To set the proper levels of safety stock you will need to understand the absolute error between your forecast and your demand.  Then you will need to determine the standard deviation of that variance.  This was discussed in episode 4 of CashFlowABC.  Please review that episode for more details and examples.


Desired service level – The amount of safety stock you carry will be directly affected by your desired service level.  For those dreamers or naïve individuals in your company a 100% service level is unrealistic.  It serves as a goal but realistically cannot be attained.  If you do have a 100% order fill rate I guarantee you have too much available stock or your forecast was overstated.  Additionally, if your product is extremely low in cost and stock outs are unacceptable to your customers then it may necessary to carry enough inventory to eliminate stock outs.  But if your product isn’t next to free then you should choose a realistic service level and put waste reduction processes in place to continually reduce your inventory levels while increasing your service level.  For those realistic individuals APICS has come up with safety factor multipliers to help set safety stock levels based on desired service levels.  The higher the service levels the higher the multiplier and safety stock.  As an example, if your standard deviation is 100 and your desired service level is 95% your safety stock would be 100 x 1.65 or a safety stock level of 165 units.  If your desired service level is 99% your safety stock multiplier is 2.33 or a safety stock level of 233.  This means for a 4% higher service level you will need to carry twice the safety stock.  This is why removing waste from your processes and lowering your inventory will save you cash.  Also, reducing your forecast error will reduce the variation and your safety stock.  Below are the service level multipliers for both the standard deviation and the MAD:


Desired Service Level %



Standard Deviation





































You can apply these tools to other variables other than demand variation.  You can apply this to variation in supply.  If you have a production process that provides variable amounts compared to the ordered amounts this can be applied to account for that variability.  Or if you have a supplier that is unreliable but provides other services that increase their value you can apply this method to that supplier to create safety stock to absorb that availability.


To quickly recap:


  • Frequency of ordering will affect safety stock in relation to the frequency and duration of the stock out
  • Variability of demand must be known so you can determine the standard variation or MAD.
  • You must decide on a service level and that level will affect your safety stock profoundly.

 Leave a Reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>