How do you prepare the safety stock?
- 2018-02-27 11:13
- Safety inventory ABC method Buffer stock method for data cube business
It's out of stock again.
Chasing the Chinese cabbage, fuck the white powder heart.
Safe stock is not right, naturally there is no sense of security.
Today, let's say how to keep the safety stock.
Safe stock setting, commonly called "ABC law". Product sales from large to small, the first 60% sales contribution for A, 30% for B, at the end of 10% to C, 631". Then the number of different inventory days is set according to the ABC class. Of course, this "631" can be changed to "721" or "532" according to the different industry, which means the same meaning.
The logic is to make the most goods in stock and make the smallest contribution. Sense: small ones, do not lose the star product. Easy communication: two months of safety stock, three times will stock oversold.
But there is also a problem: the sales of star products are large, and are prepared on a higher number of days and magnified in the same direction. once sold at a low price, it will scare people to death and sell less products. The number of minority products is not large enough. It has been covered for two months. SKU is also a headache.
So there are two advanced methods, one is the nine - Palace method, and the two is the Buffer stock method.
The ninth palace method, in addition to the consideration of ABC, introduces an axis: the prediction of the difficulty of XYZ. X class we estimate more accurate, Z class is always inaccurate, the Y class is in the middle. For example, according to the history of 6 months of forecast accuracy, take X>80%, 80%
Such a result is to reduce the sales volume of AX products with high sales volume but easy to predict, and increase the stocking of CZ products with low sales volume and difficult prediction. Meanwhile, considering the importance of products and the difficulty of prediction, make sense.
Buffer stock is the maximum value of the past accumulated oversold as buffer stock. Compared with the ABC method, it is the abandoning the days to take the units and the parameter is the count.
The following is the prediction of the forecast triangle. From top to bottom, it will be forecast from October to July next year, and rolling forecast will start from that month. This product is order to supply=1mths, block horizon=1mths, supply lead-time=2mths, 1+1+2=4, and we have to look at the deviation between the cumulative predicted value and the actual value for 4 months.
That is, the order under 10 months is the supply in November. Because block horizon (=1mths), it can only move supply from December, and lead-time=2mths will sell it in February. So October is expected to be a forecast of 10~1 months' cumulative sales.
In the orange frame, the cumulative forecast for 10~1 months is 7282, and the blue box is 11178 in 10~1 months. The two is gap 3896. This cumulative deviation, each month, will have one, and the maximum value of the past 12 months is selected as buffer stock. So in the gap line wher the 3896 is located, find the maximum value of 6096.
According to past experience, the cumulative oversold 4 months, death is 6096, then the preparation of safety stock 6096 is good.
You may have thought that the value of this buffer stock method may be very high and may be 0 at a low price. What do you do? We can set up an upper and lower limit, and make a limit according to the firm's facts.
For example, the lowest 15 days in the picture, the highest 60 days. According to the historical sales volume, we calculate the lower limit of the 1172 and the upper limit of the 4686, and finally get the upper limit of 4686 because of 6096 branches above the upper limit.
Buffer stock method, one is the difference between the total and oversold selling low, but has not reduced the prediction in safety inventory low and sell products; the two is to reduce the ABC of large amount of product delivery and the superimposed effect of small amount of products less stocking. Manual calculation is a bit troublesome, but the template is well maintained and easy to copy. In SAP, this module can be automatically updated for parameters, and we need to understand its logic.
But the difficulty of communication is a natural defect of the buffer stock method.
When the goods are ferocious, the stock is still a bit worried: is my inventory structure reasonable? So I'd like to say that a good Demand Planning is a prerequisite for Supply Planning to work well. Knowing what to sell only knows what to prepare.
Are the reserve stocks all parameter calculation and formula generation? No.
An experienced supply planner and the ability that AI can't replac easily, because he knows and judges his own product:
Which products are cheaper: there have been national stockpile of natural demand, the price of a place is?
Which products are not able to help A Dou: money to do sales promotion is also limited?
Which product gimmicks big, the effect is small: the consumer only wants one time to have and does not buy back?
......
These know that we can know the inventory risk level of different products. The risk is large. We must make assumptions and scenario analysis Scenario analysis. How much is the shortage and inventory under different scenarios? Please make financial decisions and make decisions by the general manager.
These don't know, ask more about the market and even look for the "data cube" or "business staff" in the Tmall store. Do what you can and slowly sum up. Most of the time is to eat a long cut was a wise, so supply planning is doing long, determined to do valuable.