Sku Segmentation more than ABC

Distribution requirements planning (DRP) is a popular choice among retailers & wholesalers having centralized distribution centres. Whether FMCG or General Merchandise categories, accurate forecasting is usually the starting point. While much progress has been made in this area, promotion activities and aggressive competition have a massive influence on demand creating volatility, inadvertently impacting consistency of supply. Despite relatively short lead times, avoid total stock outs can be a challenge. Many just in time replenishment modules offer opportunities for regular or daily plan runs. It favours quick reaction but whether min/max or coverage based planning approach, static uniform settings are seldom ideal.

Safety stock does provide some buffer. Statistical methods of computing safety stock weights demand variability by the risk of stock outs, given a service level, over the replenishment lead time. It is quite effective, but shouldn’t be allowed to run unconstrained. Many packages have flexibility in the safety stock output as a forward cover or absolute number in units. There are some nuances with time phased planning, but overall works quite well.

Demand variability by its nature will impact Safety Stock (SS). Unreliable supply on the other hand i.e. order fill rates, can be a source of stock outs too. This is usually a consideration for delivery strategies; frequency of replenishment, or as some DRP systems have it, the coverage period. More frequent replenishments usually mitigates supply risk.

These two concepts can be applied together by segmenting across two dimensions: ABC|XYZ. 

Using the Pareto approach for ABC: A usually makes for 70% of the total, B, 20% to 25% & C the remaining 5% or so (value or volume based; for supply planning volume is favoured).XYZ is determined on forecast error MAPE or SMAPE. As an example the following can be used for these bands: X -less than 20%, Y – 21% to 50%, Z – greater than 50%. The intersection of ABC|XYZ will help determine fulfilment parameters. 

Illustration below: (Safety Stock & replenishment Coverage example; in Days forward cover of demand):

• High Volume items with good/ reasonable forecast accuracy (AX| AY): NO need for excess inventory (low to moderate safety stock); frequent replenishment therefore coverage set below 5D.
• Low Volume items with poor forecast accuracy (CY| CZ): Items deemed as non-strategic. Set higher safety stock (cap with max); less frequent replenishment – coverage set above 7D.

Potential benefits:

- Working capital by having stock distributed proportionally on contribution.

- Allows Planners to focus their time managing exceptions on core items.

- Can inform distribution centre planning, workforce scheduling &

- Has the flexibility to accommodate space constraints in store.