Segmenting your inventory.
Inventory of finished goods and purchased items is not just items on pallets and shelves, but one of the major drivers affecting the effectiveness and profitability of your business.My last blog opened a discussion about getting to know your inventory and how it behaves.
Getting to know anything requires you to break it down to understandable elements and inventory is no different. The approach is called segmenting inventory and the most common approach is ABC analysis. The basic ABC is limited to just three segments (or categories), This means that you will treat all inventory items (SKU) within a category the same way, even though the pattern of demand for items with similar total sales or usage will vary. This is too broad a measure to enable proper management.
The ABC analysis requires an extension, so there is a greater number of categories and therefore inventory management processes to be used that better reflect the behaviour of items. An approach is the coefficient of variation for management (CoVM), developed by Tom Rafferty of Supply Chain STO. The basis of this analysis is to use the variability of sales for each SKU as the guiding measure.
The CoVM model
The CoV for a finished goods SKU is calculated by dividing the standard deviation of weekly sales (or 13 x 4 week periods) over at least the past twelve months, by the mean of the SKU’s sales. The resulting CoV for each SKU is placed into one of five classes that reflect its variability; the lower the CoV, the more consistent are sales for the SKU and conversely, the higher the CoV, the greater the variability of sales.
You are now in a position to place SKUs into a matrix table. The basic approach is that each SKU will be grouped into either: A (about 5 percent of all SKUs), B (about 15 percent), C (about 30 percent) and D (about 50 percent). The number of groups can be increased to reflect specific situations of your business or products. However, for each group the SKUs will be placed into one of five classes representing CoV ranges. ‘Class a’ is less than 0.25, ‘class b’ is 0.25 – 0.5, ‘class c’ is 0.5 – 0.75, ‘class d’ is 0.75 – 1.0 and ‘class e’ is greater than 1. In the base model, you now have 20 active cells, where for example, cells Aa, Ba Ca and Da have similar sales patterns and therefore can be managed in a similar manner.
This is not a set and forget process; if you experience cyclical demand patterns or have rising or falling sales, you should recalculate your CoV for all SKU at least twice per year and preferably quarterly.
Grouping the cells into categories and placing a management policy on each category will be the topic of my next blog.