Inventory optimization

Inventory optimization refers to the techniques used by businesses to improve their oversight, control and management of inventory size and location across their extended supply network.¹ It has been observed within operations research that "every company has the challenge of matching its supply volume to customer demand. How well the company manages this challenge has a major impact on its profitability."²

In contrast to the traditional "binge and purge" inventory cycle in which companies over-purchase product to prepare for possible demand spikes and then discard extra product, inventory optimization seeks to more efficiently match supply to expected customer demand.³ American Productivity and Quality Center (APQC) Open Standards data shows that the median company carries an inventory of 10.6 percent of annual revenues as of 2011. The typical cost of carrying inventory is at least 10.0 percent of the inventory value. So the median company spends over 1 percent of revenues carrying inventory, although for some companies the number is much higher.⁴

Also, the amount of inventory held has a major impact on available cash. With working capital at a premium, it is important for companies to keep inventory levels as low as possible and to sell inventory as quickly as possible.⁵ When Wall Street analysts look at a company's performance to make earnings forecasts and buy and sell recommendations, inventory is always one of the top factors they consider.⁶ Studies have shown a 77% correlation between overall manufacturing profitability and inventory turns.⁷

The challenge of managing inventory is increased by the "Long Tail" phenomenon which is causing a greater percentage of total sales for many companies to come from a large number of products, each with low sales frequency.⁶ Shorter and more frequent product cycles which are required to meet the needs of more sophisticated markets create the need to manage supply chains containing more products and parts.⁵ Hence, businesses need to understand how this affects their inventory and how they can seize the opportunities presented by such products.⁸

At the same time, planning frequencies and time-buckets are moving from monthly/weekly to daily and the number of managed stocking locations from dozens in distribution centers to hundreds or thousands at the points of sale (POS). This leads to a large number of time series with a high level of demand volatility.⁹ This explains one of the main challenges in managing modern supply chains, the so-called "bullwhip effect", which often causes small changes in actual demand to cause a much larger change in perceived demand, which in turn can mislead companies to make bigger changes in inventory than are really necessary.¹⁰

Without inventory optimization, companies commonly set inventory targets using rules of thumb or single stage calculations. Rules of thumb normally involve setting a number of days of supply as a coverage target. Single stage calculations look at a single item in a single location and calculate the amount of inventory required to meet demand.¹¹

Inventory optimization models can be either deterministic—with every set of variable states uniquely determined by the parameters in the model – or stochastic—with variable states described by probability distributions.¹² Stochastic optimization takes supply uncertainty into account that, for example, 6 percent of orders from an overseas supplier are 1–3 days late, 1 percent are 4–6 days late, 5 percent are 7–14 days late and 8 percent are more than 14 days late.¹³

Stochastic optimization also accounts for demand volatility which is a top priority among the challenges faced by supply chain professionals.¹⁴ For example, management predicts a 65 percent probability of selling 500 units, a 20 percent probability of selling 400 units and a 15 percent probability of selling 600 units. High service levels can be achieved with cost overruns, excessive inventory and firefighting, but higher profitability can be achieved by understanding the sources of volatility and planning appropriately. The result is a better understanding of the inventory requirements than with a deterministic approach.¹⁵

Single-echelon location problems are single-type problems such that either the material flow coming out or the material flow entering the facilities to be located is negligible. In multiple-echelon problems, both inbound and outbound commodities are relevant. This is the case, for example, when distribution centers (DCs) have to be located taking into account both the transportation cost from plants to DCs and the transportation cost from DCs to customers. In multiple-echelon problems, constraints aiming at balancing inbound and outbound flows have to be considered.¹⁶

A sequential single-echelon approach forecasts demand and determines required inventory for each echelon separately. Multi-echelon inventory optimization determines the correct levels of inventory across the network based on demand variability at the various nodes and the performance (lead time, delays, service level) at the higher echelons.¹⁷

Multi-echelon inventory optimization looks at inventory levels holistically across the supply chain while taking into account the impact of inventories at any given level or echelon on other echelons. For example, if the product sold in a retailer's outlet is received from one of its distribution centers, the distribution center represents one echelon of the supply chain and the outlet another one. It should be clear that the amount of stock needed at the outlets is a function of the service received from the distribution center. The better the service that is provided upstream, the smaller the protection that is needed downstream. The goal of multi-echelon inventory optimization is to continually update and optimize safety stock levels across all of these echelons.⁶

Multi-echelon inventory optimization represents a "state of the art" approach to optimize inventory across the end to end supply chain. Modeling multiple stages allows other types of inventory, including cycle stock and prebuild along with safety stock due to time phased demands, to be more accurately predicted.¹⁸ As part of inventory optimization, supplier performance, customer service and internal asset metrics should be continuously monitored to enable continuous improvement.¹⁹

Scheuffele and Kulshreshtha refer to inventory optimization engines or IO engines, whose function is to analyze inventory data using a holistic approach across the supply network. They note growing interest in their use and application in specific inventory fields, such as plant operations, assembly lines, and within transportation.^1: 4–7

A resilient, resurgent, sustainable and reliable supply chain creates opportunities ²⁰. Even in a world shaped by global uncertainty, some businesses are not just adapting, they are making commercial gains. These frontrunners treat resilience not as damage control, but as a strategic advantage to gain global leverage.

As consumer demand fluctuates and market conditions evolve, static inventory management approaches can no longer support the rapid and unpredictable nature of online shopping and this is where dynamic inventory steps up ²¹. Dynamic inventory replenishment strategies are very important for maintaining operational efficiency and meeting customer expectations in e-commerce fulfillment.

Over the past years, China's e-commerce industry has experienced exponential growth due to technological advancements, capitalisation on digital infrastructure, and shifting consumer behaviors ²². The Chinese e-commerce economy presents a dominant situation on trading online in the world, with over half of the retail goods sold online. This position highly outperforms other major economies, such as the United Kingdom (36%), South Korea (32%), and Denmark (20%). This rapid expansion of China's e-commerce industry demands a robust framework in supply chain and inventory management to meet the demands of a highly competitive and fast-paced market. Unlike traditional retail, e-commerce platforms require great flexibility on goods, which underscores the critical role of optimized inventory management.

For instance, among the models and algorithms Amazon, Alibaba and Shein have created are deep learning algorithms for demand forecasting ²³, inventory management models based on simulation optimization, pricing optimization for promotions, and markdown optimization for product suggestions. Over the past three years, these models and algorithms have been applied to the omnichannel retail infrastructure of many retail business entities, resulting in yearly cost reductions and profit increases of over $100 million. The e-commerce market is one of the most complex but competitive markets in the world, but in these ages of competition, inventory optimization engines make all the complex works look like a walk over. As these advanced software that capitalise on artificial intelligence, machine learning, and complex algorithms to calculate and predict ideal inventory levels across a supply chain just by analyzing demand variability, lead times, and costs, these tools balance high product availability with minimal working, production or logistical capital.

Companies have achieved financial benefits by employing inventory optimization. A study by IDC Manufacturing Insights found that many organizations that utilized inventory optimization reduced inventory levels by up to 25 percent in one year and enjoyed a discounted cash flow above 50 percent in less than two years.⁵ For example:

• Electrocomponents, a United Kingdom-based world's largest distributor of electronics and maintenance products, increased profits by £36 million by using inventory optimization to achieve higher service levels while reducing inventory.²⁴
• Castrol has used inventory optimization to reduce finished goods inventory by an average of 35 percent in two years while increasing service levels (defined as line fill rates) by 9 percent.²⁵
• Smiths Medical, a division of Smiths Group, used inventory optimization to better address demand volatility and supply variability, thus reducing the risk of both understocks and overstocks while smoothing out manufacturing cycles.²⁶