The current economic slowdown has sent many companies into a tailspin as they look for ways to cut costs and improve productivity without increasing resources. Indeed, as many businesses have been forced into bankruptcy the pressure has never been greater to closely examine internal operations, monitor performance, and find hidden inefficiencies throughout the supply chain.
It is during times such as these that marketplace conditions offer the perfect opportunity for companies to affect business change and ensure objectives are being met. A number of methods, tools and techniques are available for those wishing to pinpoint where problems may exist in their enterprise, as well as outside the four walls. By establishing performance measurements or metrics throughout the supply chain, companies are more likely to reach overall corporate goals and business strategies.
Measuring performance is not a new concept, but the key players, critical issues and data complexity involved has altered dramatically how companies approach supply-chain metrics. The traditional challenge of blending operational and strategic goals in a comprehensive enterprise-wide business strategy has been complicated further by the universal reliance on operating within independent functional silos. Plus, the inter-enterprise nature of supply-chain management has prompted many organizations to find ways and tools to track and measure trading partner performance.
"People usually find a way to hit their numbers even if they do it different than the way it was intended."
Common methods for measuring critical performance issues include establishing key performance indicators (KPIs), balanced scorecards, and supply-chain metrics. "The classic approach is to say, if you want more of something, measure it, but at the same time, companies must be careful what they measure," cautions Kent McSparran, for Boston-based Keane, an IT consulting firm. "Unintended behavior from a specific metric is a common problem. In many cases, metrics result in unplanned consequences that may cause short-term successes, but don't solve the problem."
For example, one Keane client wanted to move inventory faster throughout the supply-chain pipeline, and created a metric to reach that goal. While internal results were good and the company celebrated its improvements, in reality, it actually pushed an inventory cushion back onto its distributors. "Inventory moved faster through their own internal pipeline, but in the end, the cost flowed back to them," says McSparran.
Another common mistake is to choose only one metric, such as creating the "perfect order" or "perfect shipment," which needs more measurements surrounding it for counterbalance. A perfect order usually means ensuring every order is 100 percent accurate and delivered on time. "The concept of the perfect order is a good metric, but used alone, it is too one-sided," says McSparran. "The metric is particularly valuable because it is cross-functional and does not discriminate. If there is a breakdown in this metric, it could be due to errors in any department or with any person."
There is a greater opportunity for errors in warehousing or distribution operations that work in batch mode, says John Hill, a principal for eSync International, a consulting and systems integration firm. "If a receiver on a dock doesn't input information in a system for four to five hours, a number of events may occur during that time frame that could cause more problems downstream. If product is put away in storage but not recorded, and someone picks that material for an order, all processes are now out of sync. It is important to ensure all events are captured and recorded in a timely manner."
Games We Play
Although tying compensation directly to performance metrics is a common practice, it often leads to manipulation, particularly in large corporations, says McSparran. "People usually find a way to hit their numbers even if they do it different than the way it was intended. A boardroom executive typically is not involved at the execution level, and would not see how processes may be adjusted by workers to meet outlined goals." Also, once compensation is involved, managers lose their flexibility to make changes to the metrics that could jeopardize employee benefits.
Since many performance measurement systems are functionally focused, each area sets its metrics in isolation from those of others, which often leads to functional silos and conflicting goals. Keane, which works with manufacturers, distributors and retailers, emphasizes the importance of measuring performance across silos, and stresses optimization and supply-chain synchronization as one integrated process. "Most companies still work within departmental lines," says McSparran. "For example, they think of order fulfillment only as a warehouse management function or order accuracy as a pure customer service function."
Conflicts easily can occur when local goals are pitted against global optimization. "A specific department may be performing at its best possible measure, but its individual objectives may clash with corporate level goals," says Alec Elmore, director of professional services for Prescient Systems of Fort Washington, Pa. "For example, problems typically occur between production and sales because their objectives differ widely. Production wants to manufacture in large amounts to ensure inventory, but salespeople don't want too much product that they can't sell."
A good first step in assembling metrics is to develop a corporate scorecard that defines how a company will manage performance measurement from an operational perspective. Elmore advises organizing specific metrics within scorecards by using categories, such as financial or operational, that might list freight costs or actual versus planned work. "Getting balance is key. Companies may want to target lower inventory, but the trick is how to achieve or maintain a 99.9 percent customer service level, while also keeping inventories down."
Companies can map their internal processes to better control, measure and predict supply-chain dynamics by using the Supply Chain Council's supply-chain operations reference model, or SCOR. The model includes four basic layers with varying degrees of granularity. The model's metrics are one of its most popular features, as companies choose which ones are important for their own operations. "The manpower it takes to implement the SCOR model can be overwhelming," notes Elmore. "It requires a great deal of time, commitment and maintenance to ensure systems and people are extracting the right data, and applying it to the model for continuous measurements."
Reduction in demand variation is the driving metric that governs most supply chain activities at Hubbell Inc., an Orange, Conn.-based manufacturer of wiring and other electrical devices. The company combines a variation methodology with data from its Prescient Systems' demand planning application to meet its goal of getting supply to equal demand. Demand variation is measured by days of inventory on hand of each SKU. A standard deviation is described in terms of units, such as a variation of plus or minus 10,000 for one week.
"Our goal is to reduce inventory and meet customer service standards by forming a tightly linked supply chain," says Lou Pizzoli, director of domestic operations for Hubbell. "And the way to do this is by better understanding demand variation, developing policies and procedures to react quickly to changing demands. As a result, we expect to reduce lead times from a total cycle time standpoint." Hubbell has made significant strides in reducing safety stock by meeting metrics for in-stock percentage line fill, and order fill rates.
Hubbell's methodology is to improve quality in statistical terms to reach a reliability metric of 99.9 percent. "We measure speed, inventory and service using the same metrics throughout each plant, work center and product line," says Pizzoli. "Since we can get down to the SKU level, we can determine why we missed a schedule or why we have inventory."
If lead times are too long, demand may decrease.
The e-SCOR system from Gensym, Cambridge, Mass., is one of only a few software solutions specifically designed around the SCOR model, providing modeling, simulation, and decision support for supply chains. The e-SCOR solution standardizes the way supply chains are described, builds models to simulate their operations, and tracks SCOR's array of 200 metrics. "Inventory control is a function that everyone performs, but one company may perform it by manually sending out someone to count product, while another may use an electronic database," says Mark Whitworth, vice president of sales for e-SCOR.
Similar to SCOR, the e-SCOR system is hierarchical, consisting of three levels of need. The first level allows users to describe supply-chain players, and their relationships with one another. Level two outlines the operations strategy of each player, taking into account the type of product and manufacturing mode such as make-to-stock or make-to-order. Level three maps the physical and virtual elements of the SCOR model, describing tasks used to carry out management functions, such as how to take an order, and how to manage inventory.
After needs are identified, company-specific practices are outlined in level four, including operational characteristics such as shipping times and costs, batch sizes and production times. Most important, users specify instances of variability in the supply chain, including order frequency, or raw material lead times, for instance. "If average demand for a certain product is 200 items, but could be as high as 600 or as low as 75, the model allows users to map out the distribution curve," says Whitworth. "Spreadsheets can be used for mapping, but they don't model uncertainty or variability, and can't show behavior of a supply chain over time."
As a final step, an event-driven simulation is run that shows how a given supply chain behaves over time and gives users opportunities to develop strategies when complications occur. "It is much faster than real time because users can flow a year-long strategy, highlighting the effects of demand fluctuations," says Whitworth. For instance, by seeing a spike in demand on a simulation, and low safety stocks, they know they need to make more product. Or, they can see if lead times are too long, which would cause a decrease in demand, and a potentially lost sale.
Industri-Matematik International, a supplier of supply-chain management software based in Mt. Laurel, N.J., has introduced its own value improvement analysis process for achieving metrics. The process includes 23 benefit types and three value areas that users track: reduced working capital, increased gross profit, and lower operating costs. The goal is to measure profitability and service on each order line and aggregate data to get an accurate vision of the total enterprise. "The building blocks of the analysis are the order lines," says Ulf Casten Carlberg, director of product strategy. "They contain information on gross profits, net sales value, cost of goods sold, and most importantly, revenue leakage."
Revenue leakage - one of the 23 benefit types - consists of rebates, promotions, sales and special pricing agreements that could cause a potential increase in gross profits. As the next step in the value improvement process, a strategic value analyzer is added to the system, and works by determining profitability by supply mode and customer. "The analyzer looks at strategic sourcing issues, such as which location to pull inventory from, how to deal with special orders, and whether or not to cross-dock shipments," says Casten Carlberg. "These decisions either can add to or reduce costs, and are tightly integrated with distribution and logistics processes."
'On Time' Is Late
As a make-to-order manufacturer of interior and exterior signs, Plasti-Line not only provides customized solutions to its automotive, financial and retail clients, but also installs and maintains its products. Over the past year, the Knoxville, Tenn.-based company has narrowed in on establishing KPIs across its four plants. In the process of implementing processes for achieving metrics, managers made some surprising discoveries regarding its planning and shipping operations.
Hoping to improve upon 70 percent on-time shipping performance, management soon discovered that, in actuality, the figure was much lower because a four-day cushion was included. "If an order was due to ship on Monday, and we actually shipped on Friday, this was considered 'on time,'" explains Nathan Bretscher, director of corporate process and information improvement. "It highlighted how poor we were really performing. As a result, we shortened our planning cycle. Instead of planning in weekly buckets, we moved to daily plans, and began to look further out into the future. Now we measure on-time performance to the day and there is no four-day cushion."
Investigating deeper, Bretscher found 80 percent of orders were recorded as 'shipped' on Friday and Saturday, however Plasti-Line's carriers do not make weekend shipments. "If employees missed a shipment on Friday, they would come in to work on Saturday. But, since shipments don't go out on Saturday, product would sit on trucks until Monday. This not only contributed to a significant amount of overtime pay, but also inaccurate data."
Besides following key internal metrics such as bottom line variances, Plasti-Line also focuses beyond the four walls by measuring performance of its installers. The company uses what it calls an erector scorecard, which measures on-time installation, warranties and damages incurred at customer sites. "Installers can go online to see how they are performing compared to our standards," says Bretscher. "Then we will notice if there are any visible patterns to the results we receive, such as if damage is incurred to signs more often in a particular part of the county or with a specific installer." The company uses the Virtual Production Engine from SynQuest, an Atlanta-based supplier of supply-chain planning systems, for enhanced collaboration with suppliers and customers.
Extending metrics to an inter-enterprise level is a monumental undertaking for most companies that borders on the impractical, says McSparran from Keane. "It's more of an idealistic concept because rarely do the interests of numerous trading partners align together. Most companies have a big enough challenge to meet performance measures within their own four walls."
Taking an opposing view, Bridgepoint, a Cary, N.C.-based supplier of collaborative supply-chain solutions, believes there is inherent value in striving for inter-enterprise metrics. Offering a network business information exchange, plus analytical tools and trend analysis, the company looks at establishing KPIs from both an operational and strategic perspective. BridgePoint's software gathers data from trading partners and translates it into business knowledge that will allow companies to increase their operational performance. "Our mission is to pull inventory and cycle time out of the supply chain, and identify areas to be improved," says Richard Lower, director of marketing and product strategy.
The strategic process includes reports and analysis in the form of trading partner scorecards that measure allocation and service level contracts at various levels, such as a trade lane perspective, or by specific products. "Users can identify a set of carriers that are not performing well in a certain route or see which ones perform the best in each trade lane," says Lower.
"Our mission is to pull inventory and cycle time out of the supply chain, and identify areas to be improved."
Also included in trading partner metrics is their ability to exchange quality data, considering timeliness, relevance, accuracy, completeness and data reconciliation. The process begins with trading partners defining specific events, which are translated into standard definitions. Lower gives the following example: "A customer with five ocean carriers sends an event called an 'arrive,' but the term is used five different ways. One carrier might describe an 'arrived' shipment by the action of a container being unloaded, while another carrier describes it as a container coming into port. Without this data quality, performance measurement is useless."
BridgePoint works as an application service provider, outsourcing its software to users, and providing access via the internet. Using the software's network model and working from one knowledge base, companies can view data to monitor their performance. The system extracts data feeds from all supply-chain members and aggregates it, running a solution to all partners or organizations involved in product movement. "We identify key points of delay and failure, and present results in financial terms, which enables customers to improve cycle time and reduce inventory," notes Lower.
Operational performance statistics are posted on the network, allowing all trading partners to view their performance in relation to others. "The system expresses performance in terms of service level agreements," says Lower. "Instead of saying one carrier is 50 percent lower than another, it says a carrier is 80 percent on goal with our preset agreement. Manufacturers or distributors can use this information to leverage performance of all trading partners. They could say, 'You are performing at this rate, but we want you at this rate.' Plus, they can see others are doing it, so they know it can be done."
Timely, incisive articles delivered directly to your inbox.