# Managing Inventory - Principles of Management

Inventories are material inputs into a production process and the finished products that an organization holds in anticipation of future sales. Although many organizations have to hold inventories, they tend to be a major cost driver in manufacturing and retailing organizations. For manufacturing enterprises, inventories take three basic forms: inventories of inputs into a production process, inventories of work in progress, and inventories of finished goods.

For an automobile manufacturer, inventories of inputs include the component parts it must buy from suppliers to make its cars; inventories of work in progress include partly assembled automobiles that are still progressing through the automobile plant; and inventories of finished goodsinclude unsold cars sitting on dealers’ lots. For a retailer, inventories include the stock ofunsold goods sitting in storage or on store shelves.

The most obvious component of inventory-driven costs is inventory holding costs .whichinclude both the capital cost of money tied up in inventory and the cost of the warehouse space required to store inventory. In addition, other costs are associated with holding excess inventory. For example, in the personal computer industry the prices of components can drop significantly over the space of a few months.

If a manufacturer purchases too much inventory of a memory chip, and the price subsequently drops by 40 percent before most of that inventory is used, componentdevaluation costs will be associated with having purchased too much product too soon. Similarly, if a manufacturer produces too much of a particular computer model and has to reduce prices to shift unsold inventory, a price reduction cost must be borne (this is the difference between the planned price and the actual price).

In the case of excess production, some products might also be returned to the manufacturer from the retailer because they were not sold ( returncosts ), and the firm might have to write off the value of some finished goods that have become obsolete ( obsolescence costs ). But there are also costs associated with having too little inventory.The inability to make goods due to shortages of component parts and finished products creates stockout costs, which are the opportunity costs associated with lost business.

The total inventory-driven costs of an enterprise are thus the sum of holding costs, component devaluation costs, price reduction costs, return costs, obsolescence costs, and stockout costs. The actual magnitude of inventory-driven costs will vary depending on the nature of the business, but they can constitute a significant percentage of the overall cost structure of an enterprise. One study of Hewlett-Packard, for example, found that before it adopted inventory reduction programs, the total inventory-driven costs in its mobile computer division amounted to 18.7 percent of revenue (the study did not measure stockout costs).

Given the importance of inventory-driven costs as a component of the overall cost structure of an enterprise, managers often devote considerable attention to formulating and implementingstrategies for reducing the amount of inventory an organization has to hold. One goal often embraced is to have inventory arrive just in time to enter a production process or be placed on a store shelf for sale.

The success of managers at doing this can be assessed by inventory turnover , which is normally measured by the number of days it takes to totally replace inventory. Different businesses will turn over inventory at different rates. Wal-Mart, which is one of the most efficient retailers when it comes to inventory management, turns over its inventory every 45 days.

Dell, which is the most efficient company in the personal computer industry, turns over its inventory every two days! At Dell the mantra is to “replace inventory with information,” meaning that the only inventory Dell wants is that in transit between suppliers and Dell’s assembly operations.

The bottom line is that if inventory can be turned over more often, inventory-driven costs will be significantly reduced. The business will have less capital tied up in inventory, it will need less space to store inventory, and it will be far less likely to suffer other inventory-driven costs.

As a result, its profitability should be significantly higher. Indeed, superior management of inventory is widely recognized as a principle source of competitive advantage at both Wal-Mart and Dell and a major reason why both of those enterprises are on the efficiency frontier in their respective industries. So how can managers increase inventory turnover and reduce inventory-driven costs?

ECONOMIC ORDER QUANTITY AND SETUP TIME

One starting point of inventory analysis is to determine how much inventory should be held. A widely used algorithm, called economic order quantity (EOQ), helps managers to do this.

EOQ is defined as follows:

EOQ =(2* D*FC)(VC* K)

where:
D = Annual demand
FC = Fixed costs of producing/procuring inventory
VC = Variable costs of inventory
K = Inventory holding costs

For illustration, suppose the manager of an automobile assembly plant has to decide how many right door panels to order from the body stamping shop at a time. The factory produces 200,000 cars a year, so total annual demand for the door panels is 200,000. The fixed cost of setting up equipment in the stamping shop to produce the panels is $10,000, and the variable cost of each panel is$50. The inventory holding costs (the costs associated with the capital tied up in the inventory and the cost of warehousing) are 20 percent of the variable costs. Then

EOQ = √(2*200, 000* 10, 000) (50*0.20) *20, 000

This algorithm tells the manager that the economical batch size is 20,000 units. The factory consumes 200,000 right door panels a year, so she will have to reorder 10 times a year (200,000 divided by 20,000). Because there are 365 days in a year, sheshould reorder every 36.5 days (365 divided by 10). In other words, under this ordering system the inventory would turn over every 36.5 days.

We can also use the algorithm to calculate the cost of holding the inventory, which is the holding costs, K, times the variable cost, VC, times half of the EOQ. (We are assuming that the inventory is depleted at a steady rate and the new inventory arrives just as the inventory in storage is depleted, so that the average inventory in storage is half of the EOQ.) Thus the inventory holding costs are

Inventory holding costs =K * VC *1⁄2 (EOQ)
=20% *$50 *1⁄2 (20,000 units) =$100,000

So far so good! However, if the goal of the firm is to maximize profitability, the manager should see the $100,000 in inventory holding costs as something to be reduced. This was exactly the issue that confronted OhnoTaiichi, the engineer at Toyota who pioneered the development of flexible production systems in the company. Ohno’s solution was to reduce the fixed costs of setting up equipment to produce inventory (FC). As can be seen from Table, which shows the EOQ and inventory holding costs using different assumptions regarding fixed costs, the impact of this can be quite dramatic. If the fixed cost can be reduced from$10,000 to just $1,000, for example, the EOQ falls to 6,325 units, and inventory holding costs decline to$31,623. As discussed earlier, at Toyota Ohno and his team reduced setup times (which drive fixed costs) from a day down to minutes.

The effect was to boost inventory turnover and reduce inventory holding costs. The last row in Table is a good analogy for what Ohno achieved by his development of flexible manufacturing technology: The fixed costs associated with setting up machinery were reduced to almost zero (just $10 in our example), EOQ batch size fell to about a day’s worth of inventory, and holding costs became trivial ($3,162 in our example). As a result, Toyota’s profitability increased.

JUST-IN-TIME INVENTORY SYSTEMS

By pioneering attempts to reduce setup times, Ohno not only invented a flexible manufacturing technology, but also created the concept of just-in-time inventory. When the fixed costs associated with setting up equipment to produce a component part can be reduced to almost zero, the parts can be produced economically in very small batches and enter the production process almost immediately—in essence, just in time for them to be used.

To coordinate theflow of materials within Toyota, Ohno developed another process innovation known as the kanban system. Under the kanban system, component parts were delivered to the assembly line in containers. As each container was emptied, it was sent back to the previous step in the manufacturing process. This became the signal to make more parts.

At Toyota the just-in-time system was originally developed in-house, but over time it was also extended to embrace most of the company’s suppliers. Toyota helped its suppliers adopt flexible manufacturing technologies and the kanban system. Many of these suppliers located their plants adjacent to Toyota’s assembly operations to reduce the amount of inventory in transit as much as possible. By doing this Toyota was able to extend the gains from in-house inventory reductions to its entire supply chain, which reduced the costs it had to pay for component parts from suppliers and therefore its overall cost structure.

Toyota’s just-in-time model has now been replicated in many organizations and, as at Toyota, has often been extended to even those with globally dispersed supply chains. In a 21st-century variant of the kanban system, Dell Computer electronically shares information regarding the orders it receives through its Web site on a real-time basis with its suppliers. This helps Dell’s suppliers optimize their own production schedules and inventory, producing just what Dell needs. The flow of parts in transit, whether via ship, plane, or truck, is tightly controlled and synchronized so parts arrive at Dell’s assembly plants just in time to enter the manufacturing process.

As at Dell, modern information systems and flexible manufacturing technologies have made it easier for other organizations to adopt just-in-time inventory systems. Flexible technologies mean that more goods can now be economically produced in smaller lot sizes. Bar code scanning of inventory, along with Internet-enabled data links between a firm, its suppliers, and transportation providers shipping inventory, have made it possible to tightly control the flow of inventory, which reduces inventory-driven costs.

Just-in-time systems are also important for retailers. If inventory arrives at a store just in time to be placed on the shelf for sale, the store does not have to devote space to holding inventory (reducing inventory holding costs), which means more space can be devoted to displaying goods for sale. Moreover, just-in-time delivery of goods means the retailer is less likely to be left with inventory that it cannot sell or has to mark down, thereby avoiding price reduction and obsolescence costs.

A final tangible benefit associated with just-in-time systems is that defective parts show up in the production process soon after they have been made. This makes it easier to identify and fix the source of the defect than if the parts had been stored in a warehouse for an extended period. In effect, just-in-time inventory systems help managers improve product quality by eliminating defects from the production process.

The biggest drawback with just-in-time systems is that they eliminate inventory buffers. Buffer stocks of inventory are inventories held for some unexpected contingency—for example, if the equipment in a supplier’s plant breaks down and has to be replaced. If no buffer stock is held in reserve, the entire assembly operation could stop while the firm waited for its supplier to fix equipment.

As insurance against such contingencies, some organizations keep buffer stocks of inventory on hand—perhaps a few days’ more supply than they need. Of course doing so raises inventory-driven costs. Managers in some organizations, such as Dell, have decided to operate without such buffer stocks. This means that from time to time, when unanticipated events occur, they may have to scramble to solve a sudden shortage of inventory.

This happened to Dell twice in the early 2000s: once after the terrorist attacks on the United States on September 11, 2001, and once during the outbreak of the SARS virus in China during 2003. Although the total shutdown of ports and airports in the aftermath of September 11 hurt Dell, it was able to adapt to the slowdown of supplies from China by reconfiguring its supply chain, pulling supplies from ships and flying them directly to the United States.

BUILD TO ORDER AND INVENTORY

A final way of reducing inventory-driven costs is to move from a build-to-stock business model toward a build-to-order business model. In a build-to-stock model, a business makes products to stock the distribution channel in the anticipation that customers will purchase those products.

Thus General Motors builds cars to stock dealers’ lots, Hewlett- Packard builds personal computers to stock the shelves of electronic retailers, and Levi makes jeans to stock the racks of clothing retailers. In a build-to-order model a firm takes an order first and then builds the product. When Lands’ End sells customized shirts and pants over the Internet, it is adopting a build-to-order strategy—as does Dell when ittakes customer orders for personal computers over the Internet, or Boeing when it takesorders to build commercial jet aircraft.

Build-to-order systems have two big advantages over build-to-stock systems. First, because the company knows exactly how much of an item it will sell in advance of production, it can order just enough parts for orders in hand. Second, a build-to-order system eliminates the problems associated with building too large or too small an inventory of finished products.

Thus the firm does not have to mark down products to shift unsold inventory and does not suffer stockout costs. Under a build-to-order system the firm does not have to guess what demand will be, so it can closely align supply and demand and reduce its inventory-driven costs.

With these advantages, you might wonder why all firms don’t build to order. There are several reasons. First, building to order normally implies smaller production lots, which may drive up costs, particularly for products where mass production is still the best manufacturing method (such as making toothpaste).

Second, consumers may not want to order individual items, instead preferring to walk into a store and purchase what they want (nobody wants to place a special order for toothpaste). Third, it generally takes a while to build products to order, and consumers may not want to wait. (Who wants to wait two weeks for toothpaste to be delivered?)

Finally, legacy considerations may make it difficult for a business to move from a build- to-stock to a build-to-order business model. For example, Hewlett-Packard (HP) would love to run its personal computer business entirely on a build-to-order basis, as Dell does, basing production on orders coming in over the company Web site. However, HP has a vast network of dealers that currently sell its products.

If HP shifted to a Web-based build-to-order system overnight, its sales would probably slump because the increase in Web sales would not likely offset the decline in sales through traditional retail channels. Interestingly, HP is starting to move toward a build-to-order model and now sells personal computers via the Web; butit still distributes the vast majority of its personal computer products through traditionalchannels.

As with HP, many manufacturing enterprises, particularly those that make large durable products, are trying to move toward a build-to-order business model, or at least increase the proportion of their sales that come this way.

The rise of flexible manufacturing systems, with their implications for mass customization, has made build-to-order systems more economical. Several automobile companies are experimenting with a build-to-order approach in an attempt to eliminate the problems that arise from inventory overstocking or understocking. Nissan has calculated that if it could move to a build-to-order system with a short cycle time, it could reduce costs by as much as \$3,600 a vehicle.

In Germany, BMW now builds 60 percent of its cars to order, but the delivery time can be as long as two months. Toyota, too, is trying to build more cars to order. In late 2004 the automaker claimed that it was building about 11 percent of the cars it sold in the United States to order, with a build time of just 14 days.