Safety stock is set to reduce the out-of-stock problem caused by uncertainty factors. If safety stock is not set or is set too little, it will lead to the loss of potential sales opportunities. If safety stock is set too much, it will lead to inventory backlog and rising storage costs. Therefore, the calculation of safety stock is to find the best balance between supply and demand. In this way, this problem becomes a math problem to find the optimal solution. Its complexity is affected by the two core variables: uncertainty in supply and demand.
Before thinking about the calculation formula, you must also consider the following problems that will be encountered in actual scenarios: inventory backlog, supplier unreliability, product production cycle, logistics problems, sales forecast, inventory management, etc.
First, cross-border e-commerce sellers input the prepared sales data into Excel tables for simulation. The supply and demand units are measured in months (can also be weeks or days).
(-)Method 1: Max-Average formula
SS (Safety Stock) = (Max Salex Max Lead Time) – (Average Salex Average Lead Time)
= 1200×1.31-1000×1.15 = 1572-1150 = 422
This formula is relatively simple, but its disadvantage is that if the average sales volume and the average supply cycle have spike peaks, the safety stock deviation will be extremely large, resulting in a large amount of inventory backlog. To answer this question, we have to face two normal distributions: one is supply, and the other is demand. Safety stock is the standard deviation of a certain supply level during the fluctuation of these two normal distributions. We use normal distribution to predict the probability of generating a certain sales volume (in fact, it is also the probability of meeting a certain sales volume without out-of-stock). We call this probability the service rate (service rate / service level), which will affect the calculation of the safety factor. The safety factor is the inverse of the standard normal distribution of the service rate (inverse of standard normal distribution). Running the python code can realize the calculation (see Figure 6-11).
When the service rate is 90%, the safety factor is scipy.stats.norm.ppf (0.9)=1.28. It means that when the safety factor of the standard normal distribution is between [-1.28,1.28], it can cover 90% of the cases. So when the safety factor is greater than 1.28 times the standard deviation, it can provide a 90% service level and a 10% probability of out-of-stock. The larger the safety factor, the higher the coverage and the lower the out-of-stock rate. (Note that the safety factor does not apply to seasonal products). When modeling, the abbreviations represent the following meanings: SS is Safety Stock; Z is Safety Factor;
DSD is Demand Standard Deviation;
ALT is Average Lead Time; LTSD is Lead Time Standard Deviation; AS is Average Sale.
(II) Method 2: Normal distribution with only uncertain demand
The standard deviation of demand can be calculated using the Excel formula STDEVS as 141.42.
SS = ZxDSDxALT
=1.28×141.42×1.15≈194
(III) Method 3: Normal distribution with only uncertain supplySS=ZxASXLTSD
=1.28x1000x0.14≈179
(IV) Method 4: Normal distribution with uncertain demand and supplyFor example: the supply of umbrellas will not affect their demand, the two are independent of each other.
SS=Zx/ALTxDSD2+(ASxLTSD)2
=1.28x/1.15×141.422+(1 000×0.14)2≈264
(V) Method 5: Normal distribution with uncertain demand and supply and interdependence This is the sum of the results of Method 2 and Method 3.
SS=ZxDSDxALT +ZX ASXLTSD
=194+179=373
If the sales volume of the product is very low, you can use the Max-Average formula; if the sales volume is very high, you can use Method 2; if the supply cycle is unclear, you can consider Method 4.
One of the most appropriate ways to conduct procurement management and inventory management is to choose a suitable ERP software and formulate rules and algorithms based on the actual sales situation of the store.