Optimising inventory where to start?

Managing increased complexity during a recession
Over the past few years, supply chain and manufacturing experts have been faced with a new set of challenges, such as product proliferation, shorter timeframes for the development of new products and global competition. As a result, the complexity of supply chain and manufacturing processes has dramatically increased. Several best practices, operating models and technologies have been put forward to manage this increased complexity. However, the current risk of a global recession is a new dimension that needs to be taken into account as it further increases the complexity of manufacturing industries.

As the economic downturn looks set to deepen in 2009, a natural reaction would be to stop all investment in supply chain and manufacturing projects. Yet, supply chain and manufacturing software solutions are important to the ongoing success of improvement programmes and operational excellence. The key is to carefully select the project that can best help a company navigate through todays tough economic environment.

As Steve Banker from ARC Advisory Group suggests, Demand Planning is a good example of a software investment that could be postponed, while Inventory Optimisation is a good example of a software investment that should be pursued. In fact, the benefits associated with an inventory optimisation project are particularly pertinent during a downturn: reduction of working capital, reduction of the risk of inventory write-off and improvements in service levels.

Where should inventory be optimised?
When we talk about inventory optimisation projects and inventory optimisation tools, one of the challenges is that the term inventory optimisation is used to identify different needs and different tools. Today, most companies claim that they optimise inventory and most software companies claim that they provide inventory optimisation tools. The reality is - of course - very different.

Distribution-focused inventory optimisation tools can be used to define the flow of products in the distribution network and the safety stock of each node of the network to achieve required service levels. Strategic inventory optimisation tools can be used to define the push-pull boundary in the supply chain network that includes raw material buffers, plants producing intermediate products, plants producing finished goods and distribution centres. Finding the optimal product mix to achieve a global service level target is a third form of inventory optimisation that may again require a different type of tool and optimisation model.

Dynamic safety stock in manufacturing
Many chemical, pharmaceutical and Consumer Packaged Goods (CPG) companies place large amounts of inventories at warehouses near manufacturing sites. Part of this inventory represents cycle stock and is due to batch production: high changeover costs push manufacturers to produce in large batches, creating high inventory. But part of this inventory is safety stock that enables manufacturers to manage demand, supply and manufacturing variability.  Reducing inventory is always advantageous, and can provide competitive advantage in two cases when the firm is subject to high changeover costs, e.g. in the chemical industry, and when product shelf life or product obsolescence are business risks.

To find the appropriate tradeoff between changeover costs, inventory costs and service levels, most companies apply a two-step process.  In the first step, they compute safety stock levels and in the second step, they compute a production plan using safety stock as a constraint that the plan needs to satisfy.

Roughly speaking, safety stock is computed in the following way: for a given product that is produced on average once every two weeks and for which the company would like to achieve a 95 per cent service level, the minimal amount of product needed to be kept in stock should allow the manufacturer to cover the demand of two weeks with a probability of 95 per cent. Because this product is not always produced every two weeks - sometimes every week, sometimes every three weeks the manufacturer will also need to take into account variability in the production lead time and safety stock therefore increases.

Simple formulas exist to compute the minimal inventory to guarantee a given service level, or a given fill rate. The problem, of course, is that these safety stock levels are then fed into a production planning system that computes the production frequency of every product. These frequencies may be far from the initial assumptions, invalidating the stock targets. This scenario could then lead to maintaining the wrong amount of inventory, resulting in both higher inventory levels than necessary and stock-outs which hurt service levels. This process, where safety stock calculations are decoupled from determining production plans, leads to what we call static safety stock, that is, safety stock levels that are based on average production frequency. Our experience is that production planners are either not aware of the problem, or when they understand that average production frequency may be different from actual production frequency, they compensate for the difference by building more inventory than necessary.

A more effective strategy is to compute the safety stock necessary to manage variability based on actual time until the next production event. In such an approach, production plans and safety stocks are determined simultaneously, resulting in dynamic safety stock. The idea is that the closer a manufacturer gets to the next production event, the lower the safety stock needs to be. In this approach, assumptions on the average production lead time which might be wrong do not need to be made. Indeed, if the production plan and the safety stock are simultaneously computed, it is clear when the next production is going to occur. As a result, production variability only needs to be considered in relation to the probability of unexpected events in manufacturing (e.g. a machine breakdown).

Benefits of dynamic safety stock
There are hard and soft benefits associated with manufacturing inventory optimisation.  The hard benefits include preliminary results which show a reduction of safety stock by 25 per cent and a reduction of total inventory costs by 10 per cent. Besides improvement in working capital, reducing inventory means reducing the risk of write-offs due to obsolescence or the risk of product waste due to shelf life. In 2008, a company in the CPG industry had to destroy 8 per cent of its entire 2008 production due to shelf life problems.

Soft benefits can be just as important. Planners can now benefit from a much more precise evaluation of the risk of stock-out and therefore can make more informed decisions. In fact, not only is the safety stock optimised together with cycle stock and manufacturing costs, but it is dynamically recomputed when manual changes are made to the production plan. The ability to make more informed decisions is crucial to enabling manufacturers to improve their flexibility in the current environment.