By Mustafa Çagri Gürbüz
Inventory management policies that fail to keep pace with shifting product demand can lock up valuable working capital, especially in volatile markets. By dynamically managing inventory without compromising service levels, companies can free up these unexploited reserves of working capital.
A research project at the Zaragoza Logistics Center, Zaragoza, Spain, has developed models for achieving these goals. The work was carried out by Rajesh Kella and Christos Agrogiannis in collaboration with the global chemical company Clariant for their MIT-Zaragoza Master of Engineering in Logistics and Supply Chain Management (ZLOG) thesis. *
Companies commonly use inventory management policies that employ standard formulas for specific demand distributions and remain unchanged for several years. It’s often easier to continue with current practice, particularly where portfolios include hundreds or even thousands of products stored in hundreds of locations.
In the meantime, however, demand patterns can change, and the policy parameters can become outdated. Forecasts inaccuracies and unrealistic service level targets make it difficult to effectively balance supply with demand for each product.
The phenomenon can be particularly damaging in fast-moving businesses where product life cycles are short, or in markets that are inherently unpredictable.
Thesis sponsor Clariant faced the latter challenge. The research focused on a single business unit in South America, where economic performance in the region’s emerging countries can be erratic. In addition, warehouse replenishment processes lacked transparency and differed from one part of the region to another. Managing inventories across multiple countries required too much manual work and firefighting.
One of the research objectives was to identify the factors that influence safety stock in the region. The other key issue addressed by the researchers – and one that is central to the idea of freeing up financial resources through improved inventory management – is achieving the right balance between net working capital (NWC) and speed of response.
The tradeoff stems from the fact that lower inventory levels are known to liberate working capital, but at the same time damage speed of response. If stocks are cut too much, service levels can decline to a point where the company risks alienating or losing customers.
Companies try to employ inventory management practices that minimize the amount of working capital they commit to stored product while maximizing responsiveness. But this is very difficult when these practices have stagnated and are out of step with market demand.
A solution is to carefully monitor the fluctuations in demand over time, and make the inventory management policies dynamic so they can flex with the market’s ups and downs.
In the business-to-business markets served by Clariant, customer orders are either fulfilled from available stock – the faster option – or from the next incoming shipment. Speed of response is measured as the percentage of customer orders that are fulfilled on time.
The research project encompassed 16 affiliate warehouses in eight South American countries. This network of facilities, fed by a global distribution center in Europe, supplies the region.
The factors that impact NWC and responsiveness are forecast accuracy, inventory level, inventory in transit, target service level and lead time. Interactions between these factors influence overall performance. For example, an improved forecasting system increases forecast accuracy which reduces uncertainty and the need to hold high levels of safety stock.
These factors are driven by product segmentation strategy, inventory policy, and network design, and the researchers looked at each of these drivers to determine how they can be shaped to meet the project’s goals.
For example, in order to differentiate the replenishment process for every SKU, the researchers reviewed two key considerations: the sales performance and demand predictability of each SKU. This means that the segmentation was not based solely on the sales performance as in the standard ABC classification, but also on demand variability. They segmented the products so that higher service levels were assigned to the most profitable and predictable ones. Less critical, low demand/unpredictable SKU’s were assigned lower service levels in order to reduce inventory holding costs.
Specific models were established for the critical items, rather than standard ones that are not effectively attuned to market conditions. Simulation models that mimic real-life operations were used to analyse demand distribution in detail instead of simply assuming normal distribution, and to optimize management policies. The flexibility of these models makes it possible to investigate future scenarios where constraints such as minimum order size and supplier capacity change.
The software tools used to model replenishment processes and adjust inventory management policies are commonplace in the supply chain field. And although the project focused on the business-to-business operations of a major chemical company, the modelling methods can be applied to any market.
The researchers report that dynamic segmentation strategies and inventory policies can reduce inventory levels by as much as 32% for certain products, and help strike a balance between NWC and speed of response. How frequently inventory management policies for high-performing products are modified depends on variables such as the type of product involved and the maturity of the market. But the end result should be the same – dynamic inventory management and the availability of more working capital.
*Inventory Optimization as a Business Advantage, by Rajesh Kella and Christos Agrogiannis. This research is sponsored by Luis Olavarria, Supply chain Black Belt from Clariant. The thesis advisor is Dr. Mustafa Çagri Gürbüz, Professor of Supply Chain Management at the MIT-Zaragoza International Logistics program. For more information on the thesis research contact Mustafa Çagri Gürbüz at: [email protected]