Capabilities > Business Intelligence > Inventory Analytics

Customer Success Story

Powerful analytic application helps reduce inventory levels drastically while improving service levels

Situation: Working capital in the form of inventory-on-hand had grown precipitously

Approach: Combine operations research, statistical engineering and practical "do-it-fix-it" decision-making to reduce inventory levels

Solution: A web-enabled decision-support system designed to optimize the inventory management process

Inventory Analytics

 

In manufacturing settings where profit margins are often razor thin, managing inventory continues to be one of the most challenging issues for companies seeking to stay ahead of the pack. Much time and effort is dedicated to forecasting and modeling with intent being to avoid two likely scenarios - million of dollars of working capital tied up in inventory or risking customer defections due to delivery problems caused by lack of inventory. Many who have experienced these "feast or famine" scenarios have developed home-spun strategies targeted at managing this predicament. In the last few years, Business Intelligence solutions have introduced new scientific clarity to the age-old inventory management predicament. At Confida, we use a novel approach combining scientific decision-making with common-sense to arrive at an optimal solution for managing inventory levels using a prioritization scheme.

In grappling with the issue of effectively managing inventory items, a company must typically consider whether an item can be managed and more importantly whether it is necessary to do so. The first of these factors, Feasibility, correlates to the ease with which an inventory item can be managed. The second factor is Importance, a characteristic that correlates to the urgency or criticality being able to of effectively manage the item in question.

Combining these two factors offers a practical approach to prioritizing the management of inventory items. Below figure shows the different levels of challenge a company would face when trying to manage the spectrum of inventory items.

 

Bus _intel _inventory 1 Copy

 

The items with the lowest levels of feasibility combined with the highest levels of importance naturally present the greatest challenge. In practice, however, we find that these items account for a small percentage of the overall population.

 

Overall, the temptation is to treat all items the same, develop the best general strategy possible and apply it to the entire universe of items. The Importance/Feasibility model, however, provides a prioritization scheme for each item according to an economic framework. This allows the items to be logically divided into the three separate zones illustrated above and for a customized strategy to be developed for each. Detailed analysis, assessment and prioritization of items, as depicted in the diagram below, is a critical part of Confida's approach.

 

Bus _intel _inventory 2 Copy

 

Regardless of the methodology used to develop such strategies, there is always a small portion of parts better-suited to hands-on management. Therefore, developing operationally different strategies for managing the different segments is not only the most effective use of scarce resources, but in particular, it allows for these resources to be applied where most value will be received.

 

For the majority of items that do not need hands-on management, Confida's approach is to use a multi-echelon logistical methodology with "scientific emphasis" based on reorder-point (ROP) technology to assist with inventory decision-making. ROP, as distinct from most in-place MRP production planning systems, is a more flexible approach which can handle fluctuations and variability in demand that often causes MRP systems to break down fully. Although ROP has shortcomings, they are less severe than MRP's, since it is more of a "buffer-based" system allowing for a higher margin of error and a simplistic and therefore more predictable supplier relationship. Furthermore, the real-life limitations can be generally marginalized with more of a pragmatic outlook into all required inputs and with a technology platform helping in the measurement and monitoring of the selected inputs. The bottom line as depicted in the diagram below is to reduce both the level and the variability of on-hand inventory.

 

Bus _intel _inventory 3 Copy

 

To do so, Confida employs a three step process involving analysis, assessment and software implementation. As depicted in the diagram below, the three steps use pragmatic ROP-logic with a continuous monitoring mechanism.

 

Bus _intel _inventory 4 Copy

 

The initial stage involves the input of all measurable data to determine the relevant policy outcomes (ROP, ROQ etc.). The development of usage data is broad-based, incorporating both historical and forecast data from all areas including manufacturing and support. Once policies have been developed and are being implemented, the focus should then shift to monitoring allowing management to continuously monitor the key input and output parameters (i.e., average and variability in usage and lead time, physical inventory and service levels). When any of these parameters falls outside of pre-set control limits, the inventory item affected is placed in an alert status and the key decision makers are warned of the possible trouble.

 

Confida's methodology for optimizing inventory levels thus provides a rigorous yet practical approach to tackling the age-old inventory versus service level problem. The overarching philosophy is to segment inventory items and prioritize interventions based on economic factors. Some items should be treated using exception processing, but most items should be administered using a scientific yet manageable methodology such as ROP. Determining reorder points and reorder quantities can then be algorithmically calculated for all items not requiring exception processing. The combination of a system such as Confida's Material Management Portal (MMP) that provides intelligent alerts notifying decision makers of outliers and timely intervention by humans provides the most effective and resilient inventory management system. We have seen inventory levels drop by as much as 60% without affecting service levels negatively using a minimally invasive implementation such as MMP (since it coexist with in-place ERP systems).