Our unique simulation method calculates the number of days demand that your inventory must cover to meet your service level target by simulating the way supply planning will work with the combination of demand and supply characteristics for your products.
Our simulation relies on statistical distributions of actual demand and lead time but has a very important advantage over simplistic methods - We use an iterative method to obtain these distributions which does not allow negative values or lead times which would result in deliveries being out of sequence. The results are a good representation what you will see in the long term to match the average demand and standard deviation that you measure. Looking at safety stock calculations touted on the web you should beware of two very unreliable methods:
- Safety stock = z x std dev of demand x square root of lead time. This is based on a normal distribution that can go negative and will give answers that are often 20-30 % too low resulting in disappointing service. The formula is, at best, an approximation and is not dimensionally correct. As demand accuracy reduces the result becomes less useful!
- Max - Average. The idea of this is that you calculate the maximum demand x maximum lead time and subtract average demand x average lead time. It gives an illusion of covering the worst case - we don't know if we have seen the maximums of demand or lead time that will occur in future. Never mind the conceptual problems - this method multiplies two outliers and the result is massive over stock with a high likelihood of waste. If everyone globally adopts this it will cause a catastrophic waste of precious resources!
Our simulation uses 100,000 planning periods - if you plan in weekly periods this is like using 2,000 years of experience in setting the right stock level.
The right theoretical level of stock depends on your target service level, your demand variability and the predictability of your suppliers or manufacturing processes (if you make your own products).
Demand variability is governed by how accurately you can forecast and by your lead times. If you know the demand that you have to satisfy in future periods then this also has an impact.
On the supply side of your planning process it is not just the reliability of your supplier or manufacturing plant to deliver on time but also how frequently you can replenish and how much you have to make or buy each time you replenish stock that determine the desirable stock level for your products.
Read on too get more detail on how these factors influence the optimum stock level for you...
Service Level & Demand Planning Factors
Target Service Level
Simply put this is how good you want your inventory to be at fulfilling customer's orders. The higher the Target Service Level you aim for the more safety stock you will need.
The measure of this service level is quantity fill rate for a period:
Quantity fill = Sum of quantities shipped / Sum of quantities required for sales orders as a percentage.
To ensure good service on the products most important to your business we advise using a higher target service level for these products. This "segmentation" can be made using ABC classification in which products are sorted in descending value of sales:
- A class - the products which contribute the top 80% of total sales value. (Typically this is 20% of your products according to Pareto principle.)
- B class - the next 15% of total sales value
- C class - the final 5% of total sales value
The higher the accuracy of your demand plan the lower the safety stock you need.
Assuming that your forecast is unbiased (i.e. you are not aiming to under or over forecast) Demand Accuracy can be measured by calculating the coefficient of variation for the actual demand relative to its average value.
Coefficient of variation = Standard deviation of demand / Average demand
NB Coefficient or variation is also known as the Relative Standard Deviation of the demand compared with it actual value. It is very important to measure this using the same periods used in planning to get a reliable result.
Planning Lead Time
Lead Time is always the enemy of Supply Chain professionals - the longer it is the more safety stock you need to cover for forecast inaccuracies during the lead time.
Measure planning lead time as the number of planning periods between placing a replenishment order to receiving stock.
NB If you ever have a case where average actual lead time for your supplies is different to your planning lead time then you should ensure immediate corrective action by your supplier. If that cannot be achieved the planning lead time must be amended to reflect the actual and avoid "planning to fail".
Knowing the demand that you must supply for in future periods reduces the need for safety stock - if you know the demand over the full planning lead time and your suppliers are perfectly reliable then no safety stock should be needed.
The Forecast Horizon is the period over which demand is known and can be used in place of forecast when planning.
Some or your products may have multiple demand streams - for example many customers buy small quantities regularly but a few customers buy very large quantities infrequently. An example is where you serve a local market from your warehouse but also supply a distributor in another territory.
Some of your products may be seasonal with peak demands in certain periods.
The ProfeSSional version of our product allows multiple extra demand streams (of these types) to be simulated so that the recommended safety stock is relevant to all demand streams added together.
Supplier Reliability and Supply Planning Factors
The more variable your supplier's actual time to deliver against your replenishment orders the more safety stock you need. This applies to internal suppliers in your own factory as well as to your external vendors and the other plants in your own organization that supply to you. We advise segmenting products according to the supplier or supplying process as performance will probably by common to these groups of products.
Supplier performance should be measured by calculating the standard deviation of actual lead time from placing a replenishment order to receipt of stock.
If a supplier's lack of reliability drives a need for extra safety stock then it is best to work with them to improve reliability before investing in extra stock.
Sometimes you cannot replace a replenishment order every planning period and you will need more safety stock as a result. This often occurs where your volume is insufficient to enable it to be economically shipped e.g. where you need to accumulate replenishment quantities over say on month to justify paying for a sea freight container shipment.
Replenishment Frequency is the number of periods between placing replenishment orders.
There are several situations in which you may order more than needed for each future planning period. This can reduce the amount of safety stock needed as you will be overstocked in many periods.
Your supplier may insist on your replenishment conforming to their Minimum Order Quantity (Minimum Lot Size).
Your product may only come in full containers, boxes or pallets so your replenishment order must conform to a Rounding Quantity which can cover your demand in multiple periods.
You may decide to order for multiple weeks to reduce transactions in your warehouse. This is common practice on C class products to reduce the cost of handling many small quantities of products which are a small part of your business.