Inventory optimization—a delicate balance

Inventory optimization—a delicate balance

Emily Kay

Prevost Parts faced a classic supply chain dilemma. The 1,200-employee, after-market parts division of Prevost Car Inc. carried lots of inventory, but it still missed targeted customer service requirements.

“Throughout our network, we have a high inventory level, but we were not achieving the desired service levels,” says Dave Gilbert, supply chain director with Prevost Parts, a North American parts distributor for motor coaches and transit buses based in Sainte-Claire, Quebec, Canada. “We were not hitting targets with our fast-moving ‘A’ items and had too much inventory on our ‘C’ items.”

Gilbert’s goals were clear: Reduce the capital tied up in inventory and boost overall inventory-management efforts but “still achieve desired service levels,” he says.

Optimizing stock while preserving or improving customer service levels poses a thorny challenge for which there is no simple solution. “Proper inventory management is the easiest [supply chain need] to understand but the toughest to accomplish,” says Gilbert. “It makes a lot of sense to keep the right parts in the right locations or maintain the highest service levels on ‘A’ products, but to achieve that with the sheer volume of SKUs is not that easy.”

Weak Link

Prevost Parts is hardly alone in its search to find the right balance between carrying expensive raw materials or finished goods and losing sales due to stock-outs. In fact, inventory management remains one of the weakest links in the supply chain, says Scott Elliff, president of Capital Consulting & Management Inc. (CCMI), a supply chain-focused consulting firm.

U.S. manufacturers’ and retailers’ efforts to shrink inventory stocks don’t match the substantial benefits companies are gaining from other supply chain-related activities, Elliff notes. “Most companies still can’t accurately forecast future requirements, despite the tremendous amount of data now available all along the supply chain,” he says. “Issues such as expediting, production schedule changes, and inventory write-downs generate ongoing frustration.”

Elliff notes that a recent CCMI survey of several industries’ logistics executives reveals disappointing results for organizations that purchased supply chain technology within the last three years.

“The technology was expected to help reduce costs and inventories, cut cycle times, improve forecasting, and increase flexibility and responsiveness in areas such as planning and execution, procurement, production scheduling, transportation management, and order fulfillment,” he says. Yet fewer than 20% of those polled show a “clear and favorable return on investment.”

With obvious room for improvement and with spending on the upswing as the economy recovers, the quest to cut excess inventory and enhance customer service fueled a 5% increase in supply chain management (SCM) expenditures during 2003, according to a recent report titled Enterprise Applications Outlook for 2003: The Performance-Driven Enterprise, from the business research firm AMR Research Inc.

“Companies still view SCM as vital to improving their operational performance, especially to lower costs incrementally and reduce inventory,” says the AMR report.

Some 56% of corporate IT executives at 499 companies with SCM investments express concerns about inventory and supply chain outlays. “Users are most interested in installing SCM applications to increase operational performance to reduce inventory and supply chain costs as well as improve customer fulfillment via reduced cycle times and increased customer satisfaction,” the report states.

The Right Parts in the Right Places

AMR also notes that the most popular capabilities among the 76% of firms with ongoing SCM implementations are those that further inventory-management/optimization and order-management efforts. That finding is not surprising because inventory management remains a vexing and expensive challenge.

With $30 million (Canadian) tied up in stock, Prevost had to slash inventory costs and strengthen customer fulfillment. “We need to have the right parts in the right place to support our customers [and] improve our performance,” Gilbert says.

After a two-year study of ways to augment inventory deployment and customer service and slim costs, Prevost chose Smart Software Inc.’s SmartForecasts Enterprise over six competing software products. Gilbert estimates that the solution, linked with the company’s SAP R/3 enterprise resource planning system, will help slice stores of items by 20% in two years.

Prevost’s evaluation involved a representative sample of the firm’s after-market parts inventory. SmartForecasts’ accurate forecasting capabilities let Prevost carry the least amount of inventory and hit service goals, says Gilbert. More precise forecasting, for example, lets Prevost cut lead time and more efficiently supply back-ordered parts to its 3,400 public transit and motor coach customers. “We’re more responsive to customers’ requests,” notes Gilbert.

Prevost’s high-priority products include such hard-to-locate items as windshields, brake components, and air-conditioning items. “We can’t be back-ordered on windshields,” says Gilbert, “because [our customers] can’t find them at a local garage store. It’s critical that we have high service levels on them.”

SmartForecast’s ability to manage Prevost’s irregular demand was crucial because customers purchase 70% of the company’s parts infrequently. Intermittent demand occurs when companies can identify few, if any, identifiable patterns in demand histories.

Smart Software points out that infrequent demand is hard to forecast because traditional approaches assume that inventory is “normal,” with predictable trends and seasonal patterns.

SmartForecasts, which makes no such assumptions, samples demand histories and runs thousands of simulations of likely future outcomes to build each forecast. The ability to forecast demand accurately lets Prevost optimize its inventory mix to ensure product availability and quick delivery while minimizing stocking levels and inventory costs.

SmartForecasts, which complements R/3, delivers forecasting capabilities that help Prevost accurately estimate safety stock requirements at its eight warehouses and distribution centers. Prevost loads desired service levels and lead times into the software, which then estimates inventory levels necessary to satisfy total demand over the lead time, and calculates and compares inventory-carrying costs with current costs.

Plugging such information into R/3 lets Prevost create a final replenishment plan and enhance inventory allocation, and it helps manufacturing operations produce the required allotment of parts to meet customer demand. Prevost expects a return on its $150,000 SmartForecasts investment within a year of its January 2004 deployment.

Internal Coordination

Optimizing inventory throughout the supply chain is difficult enough, but companies must first get their internal processes in order. Optimization and managing inventory means different things to different departments in an organization, notes Allan Ayers, principal with the consulting firm Ayers Management Services.

A manufacturing manager “focuses on parts and supplies, and then moving the finished goods out of the plant,” Ayers says. “The service parts manager worries about having the right part in the right place to meet repair service response times that are sometimes measured in hours rather than days. [Distributors are] squeezed between the price increases of the manufacturer and the demands for lower prices from the customer, so they are always looking at how to reduce the cost.”

BP Italia SpA in Milan, Italy, confronted just such a quandary. Sales and marketing executives wanted the global automotive after-market and lubricants manufacturer to carry enough inventory to meet every customers’ needs. Logistics administrators, however, sought to square customer service with working capital objectives, says Alessandro Tenaglia, the company’s supply chain expert.

To optimize the balance between service and inventory levels, BP centralized 29 installations in 25 countries. To improve inter-departmental communications and delineate accurate stock and service performance targets, BP connected ToolsGroup’s Distribution Planning Model software to some 10 ERP systems throughout Europe.

The software uses a statistical, “stochastic” method to model as many as several million stock-keeping units against numerous variables, such as line order frequencies, volumes, lead times, and lot sizes, according to ToolsGroup. Such algorithms gauge uncertainty in demand and supply and generate optimal inventory targets. With targets based on specific customer service levels for each stock-keeping unit, the software calculates the ideal stock for each item, which it translates into correct inventory targets. The model adapts to a range of situations, including commercial promotions, product expiration, and seasonality.

AMR notes that other vendors using similar techniques include GAINSystems Inc., SmartOps Corp., Optiant Inc., LogicTools Inc., i2 Technologies Inc., Oracle Corp., and Demantra Inc.

The approach has paid off. BP trimmed inventory worldwide by 20% in each of the first two years of operation and improved service levels by 9%, says Tenaglia. “Reducing the number of back orders gives us the ability to have a more stable, reliable production plan,” he notes, “so we’re able to really reduce the raw material stocks.”

Beyond the obvious working capital benefits, carrying less stock brought BP other advantages as well. In addition to eliminating several European distribution centers, the shift let BP increase safety stocks on its most critical items. “There was no room to introduce new items, make particular promotions, or introduce a new range of items because there was no room [in the DCs],” says Tenaglia. “We had reached the maximum.”

And service levels, which had languished at 87%, are now up to 96%. For BP, that means “96% of order lines are served in time, in full,” Tenaglia notes.

ON Semiconductor Corp. chose i2’s i2 Inventory Optimization over several “niche” products because of i2’s broader supply chain capabilities, says John Mallon, SCM director with ON, a $1.1 billion chip maker in Phoenix, Ariz.

ON completed a pilot project with 5,500 of its 15,000 stock-keeping units. The solution, which builds on the company’s existing i2 supply chain planning system, analyzed ON’s data and demonstrated that current approaches did not optimize inventory investments to improve service performance, says i2.

ON’s existing i2 solution couldn’t handle the company’s complete range of inventory issues, which involve multiple manufacturing stages. “The prior generation solution from i2 only helped with establishing inventories at a single stage of the supply chain, without considering inventory buffers up and downstream,” says Mallon. The new software can “solve the problem across multiple stages of the supply chain,” which helps ON optimize and synchronize safety stocks throughout its supply chain.

The software creates optimized targets by combining targeted service levels with supply chain cycle and transit times and demand and supply variability, says Mallon. It also lets ON run simulations of the new targets with real supply chain historic data to validate the optimized service levels.

“This is critical to validate and gain confidence in the results before going live with the new targets,” Mallon notes. The application works seamlessly with ON’s existing i2 software, which receives safety stock targets and executes the targets, says Mallon, who adds that ON plans to deploy i2 Inventory Optimization in its eight other DCs and cross-dock inventory hubs by the middle of 2004.

Good Inventory

While most companies seek to reduce capital tied up in inventory, Sherwood Bollier, president of Niagara Cutter, a tooling products manufacturer in Amherst, N.Y., cautions that not all inventory is bad. “Inventory is critical to running the business,” avers Bollier. “Many companies overlook how important inventory is.”

Niagara’s competitive industry requires the company to stock enough items to meet customer demands. The firm manufactures 24,000 products, some 10% of which account for about 85% of sales. Customer demand for other products is irregular.

Niagara had focused on items with normal demand patterns and regular turn over, which represent only 5% of the company’s inventory. Customers occasionally request items from the company’s slow-moving inventory, which is 95% of total inventory items.

To maintain high service levels while cutting total inventory, Bollier needed better forecasting and planning tools. Because the firm’s existing manufacturing system couldn’t generate forecasts, and because spreadsheets for production planning were inadequate, Niagara didn’t forecast often; when it did, results were hit or miss.

Niagara deployed SmartForecasts Enterprise to help it re-adjust as well as reduce inventories. SmartForecasts’ ability to forecast intermittent demand was key. Within six months of deploying the system at its Massachusetts plant, Niagara cut work-in-progress inventory levels by 75%. Bollier expects a 50% enterprise-wide reduction in inventory and significant increases in inventory turns.

Based on new forecasts, the company has a better grasp on its stocking needs and can reposition items when demand changes. When SmartForecasts demonstrated that future demand at the firm’s Pennsylvania plant was shifting from its uncoated line of cutting tools to its high-performance coated line, Niagara repositioned its inventory, eliminated obsolete items, and made better use of available assets.

Niagara also forecasts more frequently and with greater flexibility. The company forecasts demand monthly for some items, such as new products. Quarterly, semi-annual, and annual forecasts are sufficient for other products. Niagara, which can set customer service levels for each item, forecasts demand for key items at a 95% service level, while it sets demand forecasts for other products at lower levels.

SmartForecasts uses demand history stored on Niagara’s daly.commerce Inc.’s Application Plus back-end software. Bollier can evaluate inventory data and identify all items that customers demand sporadically, and the system forecasts those items. Bollier processes the remaining items, combines the two forecasts, and feeds the results back to Application Plus, which plans inventory requirements.

SmartForecasts “lets us look at whether we should change our inventory position,” says Bollier. “It lets you look at current inventory levels needed for specific customer service levels.” Such capabilities have helped Niagara sustain a 95% ship level over the last five years, Bollier adds.


The holy grail of inventory optimization may be to reduce inventory, improve inventory management, and maintain desired service levels. The task, however, remains daunting.

“It’s almost impossible to be the best in class in service and working capital,” says Peter Wietfeldt, consumer products practice director at the Pittiglio Rabin Todd & McGrath (PRTM) consultancy. “You can’t have the lowest inventory and the highest service levels in your industry. It’s a tradeoff.”

PRTM recently helped a leading European consumer package goods manufacturer cut inventory from 117 days to 85 days and improve order line fills from 86% to about 91%. Within a year, the company will have inventories down to 50 days.

Because the company opted to target high service levels, order fulfillment will improve to about 98%, Wietfeldt estimates. “They recognize that while there may be companies in their sector with lower inventories, if they’re better in inventories and best-in-class in service, that’s a reasonable baseline,” he says.

The bottom line for most companies is optimizing inventory to ensure customer care. “It’s very important to manage our inventory,” says Jeffrey Ramras, SCM vice president with Applied Industrial Technologies, a $1.5 billion industrial products distributor in Cleveland, Ohio. Applied operates a homegrown solution to manage 2.5 million stock-keeping units. “From a financial perspective, you want to optimize your investment.”

But the real issue is customer service. “If you’re not taking care of your customers,” Ramras says, “your inventory process is a problem.” Fs

Reducing Inventory Is Top Business Driver

Fueling Investments in Supply Chain Management

Reasons for Installing All Total Total

SCM Software Verticals Manufacturing Services

Reduce Inventory 32% 32% 32%

Reduce Supply Chain or 24% 31% 18%

Logistics Operation Costs

Reduce Cycle Times 23% 29% 18%

Reduce Supplier Costs 20% 21% 19%

Improve Customer 19% 19% 20%


Reduce Personnel Costs 19% 21% 17%

Improve Asset Utilization 17% 17% 16%

Improve Collaboration 17% 19% 15%

Between Customers/


Improve Fill Rates 16% 21% 12%

Modernize Technology 12% 12% 12%

Reasons for Installing Less $1B

SCM Software Than $1B or More

Reduce Inventory 25% 32%

Reduce Supply Chain or 15% 21%

Logistics Operation Costs

Reduce Cycle Times 15% 21%

Reduce Supplier Costs 15% 22%

Improve Customer 21% 15%


Reduce Personnel Costs 16% 16%

Improve Asset Utilization 9% 15%

Improve Collaboration 14% 17%

Between Customers/


Improve Fill Rates 15% 13%

Modernize Technology 8% 8%

(Source: AMR Research Inc.)

Emily Kay writes about technology as a principal at Choice Communications, an editorial consulting firm in Chelmsford, Mass.

Company Information

AMR Research Inc.

Boston, Mass.


Ayers Management


Inverness, Ill.


Capital Consulting &

Management Inc.

Alexandria, Va.


Demantra Inc.

Waltham, Mass.


GAINSystems Inc.

Oak Brook, Ill.


i2 Technologies Inc.

Dallas, Tex.


LogicTools Inc.

Chicago, Ill.




Optiant Inc.

Burlington, Mass.


Oracle Corp.

Redwood Shores, Calif.


Pittiglio Rabin Todd &


Weston, Mass.


SmartOps Corp.

Pittsburgh, Pa.


Smart Software Inc.

Belmont, Mass.



Boston, Mass.


COPYRIGHT 2004 Advanstar Communications, Inc.

COPYRIGHT 2004 Gale Group