4 Maturity Levels of Using Inventory Policies

Any company which maintains an inventory uses certain inventory policies. These policies govern when to order and how much to order for each stock keeping unit. And thus, these have a major impact on the inventory related performance indicators e.g., inventory turns, service levels, obsolescence etc. In my career spanning 37 years being a user, developer, and consultant in inventory management systems I have seen different ways in which companies use inventory policies. Let us call them maturity levels

·        Level 1- Using spreadsheets- These companies use their own thumb rules implemented in spreadsheets. They pull the stock and order data into these spreadsheets and perform the replenishment calculations. It is surprising that many of these companies have ERP systems with inventory policy functionalities. However, they don’t use those due to lack of necessary domain and software knowledge and spend enormous time to trigger replenishment orders which may be error-prone and sub optimal

·        Level 2- Ad hoc of inventory policies in the ERP- These companies do define and use the inventory policies available in their ERP systems or in stand-alone inventory management software. However, the selection of the policy as well as the values both are ad hoc based on human judgement. For example, someone decides whether reorder levels and fixed order quantity should be used or Min max levels. And if it is Min Max levels then what should be the minimum and maximum stocks levels. They save the human efforts in level 1, but the inventory efficiency levels are still same.

·        Level 3- Calculated inventory policy values – At this level the selection of the policy is still ad hoc. However, the policy parameters are calculated using formulas readily available in supply chain management textbooks. One approach is to extract the historical data and perform these calculations. On other hand, many good ERP systems have these calculations built in as features which can be run periodically to calculate and update inventory policy parameters.

·        Level 4- Monte Carlo simulations- Level 3 above has two limitations. One -the selection of the policy is still ad hoc. Secondly the parameter calculation formulas have many underlying assumptions which rarely hold good in real life. For example, these calculations assume normal distributions of demand and lead times which is not always the case. Then demand could be non-stationary, intermittent, and delivered in multiple lots. The calculations don’t necessarily work best under these conditions. The best practice would be to simulate these conditions using real data distribution and test what -if scenarios for various policies and their parameters.  It is also possible to build special constraints like space, shelf life of perishable products etc. to achieve optimal results under real life conditions This approach provides significant improvements over Leveel-3. It may be noted that simulation only provides the optimal policies to be implemented in the ERP. The execution of replenishment still happens within ERP.

It is common observation that an inventory performance improvement of up to 30-40% can be achieved by simply fine tuning the inventory policies in the existing software, with some expert advice.

I would love to know your experiences and discuss further if you are interested. Please feel free to DM.