In the fall 1997 inaugural issue of Supply Chain Management Review, David Anderson, Frank Britt, and Donavon Favre set forth "The Seven Principles of Supply Chain Management." The authors argued that, if applied collectively, these principles could enable supply chain partners to enhance revenue, control costs, increase asset utilization, and improve customer satisfaction. The seventh principle stated that supply chains needed to "adopt channel-spanning performance measures to gauge collective success in reaching the end user effectively and efficiently." Indeed, most business managers would agree, in theory, that such measures are needed to motivate the type of inter-company collaboration that is needed to satisfy end customers. Yet they would be quick to point out that they work in the world of reality, not theory. The challenge is to close the gap between the theoretical appeal of chain-spanning measures and the genuine concerns that complicate their actual usage.

Closing the gap requires a performance measurement system that is quite different from those applied in traditional logistics, operations, and purchasing environments. In these arenas, the focus is on measuring efficiency and effectiveness within functions and within the sub-activities of those functions. For example, in logistics, many companies measure the length of the order cycle, the cost to select an order, and the level of customer satisfaction with service representatives. Then, within the activities of warehousing and transportation, specific success factors are measured, such as average time to unload a truck, percentage of error-free order picks, and transportation cost per mile. Both the function-wide performance measures and the specific activity-focused measures are important in tracking and controlling how well the system is performing—and how effective managers are in executing their key functional responsibilities.

In supply chain management, however, the emphasis is on how well a group of companies performs in terms of creating value for the customer. The performance measurement philosophy must be altered accordingly in three important ways.

First, companies in the supply chain must work collaboratively. A true supply chain is defined by inter-company collaboration—and the supply chain performance measurement system should reflect this. Second, in the supply chain environment, both companies and individual managers must be motivated to work in collaboration with supply chain partners. Because performance measurement systems influence behavior ("you get what you measure"), the performance measurement process must be structured to provide incentives for collaborative behavior. Finally, each company in the supply chain, regardless of how far upstream it may be, needs to focus on the satisfaction and ultimate cost of serving the final customer. Indeed, the rationale for adopting a supply chain management approach is to reduce costs, increase the speed to market, and add the most value for the ultimate customer.

This article answers two questions with respect to supply chain performance measurement.

First, it answers the question, What do I measure?, by presenting a framework introduced by Robert S. Kaplan and David P. Norton called the balanced scorecard. Although the balanced scorecard is not intended to give specific performance measures for specific business contexts, it does provide an overarching template that can guide the chain-spanning metric-selection process.

Second, this article examines the question, How do I implement a chain-spanning balanced scorecard? And it addresses the obstacles that are likely to complicate the implementation process. Supply chain partners that are able to overcome these obstacles are likely to reduce waste and compress time throughout the supply chain, while enhancing their flexible response capability. And this, in turn, results in higher profit margins and return on assets, improved cash flow, and revenue growth.

The discussion unfolds in four main sections. The first section briefly defines the balanced scorecard and explains the implementation process. The second section applies the balanced scorecard framework to supply chain management. The third section discusses the implementation hurdles that complicate the usage of the balanced scorecard in a supply chain context. The final section suggests some implementation tactics that can help supply chain partners successfully implement a chain-spanning balanced scorecard.

As the name implies, the balanced scorecard is designed to achieve a balance between financial and nonfinancial performance across short-term and long-term time horizons. The scorecard focuses on performance from four different perspectives: the customer, business processes, innovation and learning, and financial results. (See Exhibit 1 on the following page.) It moderates the tendency to overemphasize financial performance by including measures related to the underlying drivers of long-run profitability—namely business process measures, innovation and learning measures, and customer-satisfaction measures.

The logic behind the four dimensions of performance is clear. The customer perspective is designed to incorporate customers' feedback regarding their perceptions of the value proposition being offered to them. The process perspective focuses on the outputs of each business process in terms of quality, time, flexibility, and cost. The innovation and learning perspective deals with measures that assess the ability to learn, evolve, grow, and prosper over the long run. Finally, the financial perspective is that all-important lagging indicator that ensures that success in the prior three dimensions of performance ultimately translates into financial results.

The balanced scorecard implementation process can be broken down into four steps.

Step 1: Formulate Strategy and Build Consensus

Though this is the most important phase of the entire project, it often is overlooked or trivialized because of the faulty assumption that everyone intuitively understands the strategy. Yet a survey conducted by CFO Magazine found that although 91 percent of chief financial officers (CFOs) had a clear understanding of their company's vision, only 71 percent of executive managers, 40 percent of middle managers, and 3 percent of line employees did. If the level of understanding is so low within individual companies, imagine how much lower it must be across multiple companies in the same supply chain. Therefore, supply chain partners need to clearly define their strategic objectives and gain a mutual understanding of where their objectives converge, and perhaps diverge, before pursuing any particular chain-spanning performance measures. In short, the supply chain strategy needs to drive the metric-selection process, which, in turn, should drive the achievement of optimal results for each supply chain partner.

Step 2: Select Measures that are Consistent with the Strategy

There is a wrong way and a right way to select measures for a balanced scorecard. The wrong way is to use a brief brainstorming session to generate a wealth of potential measures that are quickly assembled into a disconnected scorecard. A quickly assembled, disconnected scorecard will eventually pull different functional managers and supply chain partners in competing directions. The right way is to spend the extra time to ensure that the chosen measures are Harmonized, Optimal, Parsimonious, and Economical (HOPE).

Harmonized measures are connected across the scorecard's four perspectives in a manner that motivates cross-functional managers to focus on complementary objectives. For example, assume that a research and development-intensive company wants to create a balanced scorecard to support its strategy of "being the market leader in selected technologically sophisticated niche markets." Exhibit 2 (panel A) shows four harmonized measures that could support this strategic objective.

The "innovation and learning" perspective includes a measure called "time to develop next generation" that ensures that research and development (R&D) is efficiently developing the next generation of products. The business process perspective includes a measure called "actual introduction schedule vs. plan" that holds people on the manufacturing floor accountable for taking an efficiently designed product and actually translating raw-material inputs into finished goods in an efficient manner. The customer perspective includes a measure called "on-time delivery" that holds warehousing and distribution responsible for taking the hand-off from manufacturing and actually getting the product into the customer's hands in a timely manner. Importantly, these measures span the value chain from the design stage through the manufacturing and distribution stages to ensure that cross-functional managers are working together to design, manufacture, and deliver products to the end consumer efficiently. The financial perspective includes a "gross margin" measure to ensure that the company realizes favorable gross margins as a result of being the first to market.

Optimal measures control the risk of motivating employees to overachieve on one dimension of performance, while causing sub-optimal performance overall from the standpoint of the customer. Panel B of Exhibit 2 expands upon the example shown in the top half of the exhibit by including four control measures intended to motivate optimal results. The measure "percentage of sales from products less than three years old" controls the risk that R&D will overemphasize timely development of next-generation products at the expense of quality. The measure "manufacturing yields" controls the risk that manufacturing managers will overemphasize the need to have large-scale volumes of production available in a timely manner at the expense of product quality. The measure "customer perception of technological superiority" also ensures that "time" as a critical success factor does not override the need to deliver quality products to customers. The measure "profit margin" ensures that favorable performance in terms of timeliness and quality does not come through uncontrolled spending in expenses that appear below the gross margin, such as R&D, selling, and distribution.

Parsimony refers to the decision as to how many measures to include in the scorecard. Though there is no "right" number, the goal should be to create a scorecard that is complete. That means the scorecard should motivate goal-congruent performance but do so with the fewest possible measures. A measure that cannot be connected to other measures within the scorecard should be dropped. Each measure must either work in harmony with other measures or help control for the risk of sub-optimal performance. Furthermore, any measure that is highly correlated with another measure within the scorecard is a candidate for exclusion. For example, return on investment (ROI) and return on assets (ROA) are likely to be highly correlated; therefore, both measures would not be necessary. Generally speaking, fewer measures (12 to 24) are preferable to more measures (24 or more). This helps managers stay focused on the most important dimensions of performance.

Economical measures offer value by driving optimal performance that exceeds the related data-gathering costs. Some measures with great appeal in theory have prohibitive data-gathering costs. Therefore, when discussing measures that are candidates for inclusion in a scorecard, companies need to address the question of data availability. For measures deemed critically important, it may make sense to invest in the technology necessary to support the data-collection effort. In other situations, it may make sense to track a modified version of the measure that is less costly in terms of data availability

Step 3: Link and Communicate the Measures

A balanced scorecard will be useful only to the extent that it drives action at the operational level. Therefore, the strategic, high-level measures included in the balanced scorecard must be disaggregated into component measures that are understandable and actionable by employees on the front lines both within companies and across the supply chain. For example, the strategic measure "return on assets" may be meaningful to a high-level executive but not relevant or actionable to people at the grassroots level. Therefore, measures like ROA need to be broken down into specific, actionable component measures that drive the desired improvement.

Step 4: Drive Managers to Attain Desired Results

In driving managers to attain favorable results, companies need to take five specific actions:

• Establish accountability.
• Set targeted rates of improvement.
• Create action plans.
• Perform progress reviews.
• Embed the scorecard into the organization.

Establish Accountability. Obviously, a company (or a supply chain) needs to establish accountability for achieving results; otherwise, the balanced scorecard will never find its way onto anybody's "plate" of priorities. Although some of the disaggregated measures that support a balanced scorecard will be controllable by managers within a given department, others will not be. In these situations, cross-functional accountability is appropriate and "process owners" need to be created. Appropriate time frames need to be set as well, based upon each employee's scope of responsibility. Senior managers with a larger span of control should be held accountable for achieving results over a quarterly or annual time horizon. Departmental supervisors and front-line workers, on the other hand, should be held accountable for achieving results over a shorter time frame, such as a week or a month.

Set Targeted Rates of Improvement. This is an essential first step in driving actual improvement. A baseline assessment is needed to measure current performance before targeted rates of improvement can be established and eventually evaluated. These expectations must be clearly articulated and communicated. Importantly, all of the supply chain partners who will be held accountable for achieving the results must have the opportunity to provide their input to the target-setting process.

Create Action Plans. Once targeted rates of improvement have been established, action plans need to be created and linked to specific improvement goals. These plans need to be connected vertically and horizontally throughout the organization and across the supply chain. Also, each action plan should be supported by specific cost and benefit estimates as well as by specific milestone timetables.

Perform Progress Reviews. This step ensures that results are eventually achieved. The progress reviews need to be sensitive to the lead/lag relationships between various scorecard measures. They also need to focus on the ultimate outcome of revenue growth and/or cost savings within an appropriate time frame. A process orientation and a commitment to innovation are meaningless unless they eventually result in increased profitability. If this doesn't happen, the strategy, critical success factors, and chosen measures and linkages need to be re-evaluated and possibly reformulated to ensure that the desired outcome is achieved.

Embed the Scorecard. The best way to embed the balanced scorecard into an organization or supply chain is by linking it to performance evaluation and reward. This ensures that people focus on activities that produce the desired outcomes. Furthermore, the balanced scorecard needs to be linked to the resource allocation process. In other words, capital budgeting proposals should state and quantify their benefits in terms of the balanced scorecard measures. On a smaller scale, process improvement proposals also should state and quantify their benefits in balanced scorecard terms. In short, the balanced scorecard should become the language of decision-making within an organization and among supply chain partners.

The principles of supply chain management readily dovetail with the balanced scorecard framework. Exhibit 3 demonstrates the interrelationship between supply chain management and the balanced scorecard by showing the following:

1. The goals of supply chain management (waste reduction, time compression, flexible response, unit cost reduction, and so forth) can be measured via the business process perspective of the balanced scorecard.
2. The outcomes of the SCM process—namely satisfying customers (in terms of quality, time, flexibility, and value) and achieving financial results (in terms of profit margin, cash flow, revenue growth, and return on assets)—can be measured via the customer and financial perspectives of the balanced scorecard.
3. The rate of SCM improvement (in terms of product/process innovation, partnership management, information flows, and awareness of threats and substitutes) can be measured by the scorecard's innovation and learning perspective.

Exhibit 4 (on page 54) gives two examples of supply chain-oriented balanced scorecard measures. Each example includes four measures. The first example pertains to a supply chain that is focused on time as a critical success factor. The innovation and learning measure "number of shared data sets relative to total data sets" is based on the notion that sharing information across supply chain partners will enable time compression, among other benefits. For example, if manufacturers receive daily downloads of point-of-sale data, they can more quickly replenish fast-moving merchandise. If supply chain partners can learn how to overcome the barriers that impede data sharing, they will see improvement across all of the dimensions of the balanced scorecard.

The business process measure "supply chain cycle efficiency" measures the ratio of value-added time relative to the total time spent in the supply chain. For example, assume that a supply chain takes 100 days to get raw material converted to finished goods and delivered to end customers. If value-added activities are being performed on the product for a total of 10 days, the supply chain cycle efficiency measure is 10 percent (10 days/100 days). The customer measure "relative customer order response time" confirms whether the process-oriented improvement efforts in terms of time compression are delivering world-class value to end consumers. This measure could be survey-based, with end customers evaluating supply chain response time relative either to expectations or to other competitors. The measure also could be monitored through a benchmarking process whereby the supply chain partners track the response time of competing supply chains.

Finally, the financial measure "cash-to-cash cycle" captures an important financial benefit of compressing the time a product spends in the supply chain. For a single company, cash-to-cash cycle is measured by adding the average number of days that accounts receivable are outstanding and the average number of days inventory is on hand. The average number of days accounts payable are outstanding is subtracted from this total to yield cash-to-cash cycle measured in days. For a supply chain, the measure becomes somewhat more complex. The average number of days that accounts receivable are outstanding for the supply chain partner furthest downstream is added to the average number of days it takes for delivery of raw material purchased at the most upstream link to the end customer. The average number of days that accounts payable are outstanding for the supply chain partner furthest upstream is subtracted from this total to provide cash-to-cash cycle for the entire supply chain measured in days. To the extent the supply chain partners can collaborate to reduce the number of days inventory spends in the supply chain, the cash-to-cash cycle will be shortened.

The second set of illustrative measures in Exhibit 4 pertains to a supply chain that is focused on flexibility as a critical success factor. The innovation and learning measure "rate of improvement in product finalization point" is based on the premise that postponing the finalization of finished goods minimizes stockouts and markdowns. In other words, postponement enables customer orders to drive the finalization of products in real time. A supply chain that can continually push the product finalization point closer to the moment that customers actually place orders has the flexibility to meet individualized needs in a timely manner.

The process measure "number of choices relative to average response time" assesses whether the supply chain can offer variety (in terms of pallet patterns, order configurations, SKUs, and so forth) without unduly delaying the time taken to fill the order. The goal is to either increase the variety offered while, at a minimum, maintaining current response-time capability, or decrease response-time capability while maintaining the current breadth of order configurations supported. The customer measure "customer perception of flexible response" uses survey data to determine if the customers feel confident that their order requirements will be met without experiencing excessive delays. Improving the process measure "number of choices relative to average response time" should improve the customer's perception of the supply chain's ability to respond flexibly.

The financial measure "supply chain gross margin" compares the supply chain's raw material and conversion costs to the revenue realized from the end customer. A highly flexible supply chain should be able to differentiate itself from competitors—and thus be able to earn attractive gross margins. In fact, as supply chain flexibility improves (assuming other competitors cannot match the flexible response time), gross margins should improve as well. It's a good idea to disaggregate the overall supply chain gross margin into component parts that reflect each supply chain partner's gross margin. This will reveal whether supply chain partners are working collaboratively, or if one company is profiting at the expense of its "partners."

The balanced scorecard measures just discussed are nontraditional. They require a collaborative, trust-oriented approach to business management that encourages individual companies to view their success in terms of inter-organizational supply chain performance. Of course, this philosophy clashes with the longstanding history of most organizations, which have viewed their supply chain partners from an arm's length (or even adversarial) perspective. We recognize the wide gap separating the theoretical appeal of collaborative supply chain behavior from the cultural and technical reality of how organizations have historically related to one another. The challenge is to close this gap. More specifically, organizations have to overcome eight implementation hurdles to make chain-spanning balanced scorecards a reality.

1.Overcoming Mistrust

For the better part of the last century, companies maintained an arm's length relationship with their suppliers and customers. It's not surprising, then, that high levels of mistrust still exist among supply chain members. Consequently, any new approach to measuring performance across the supply chain will be met with a certain amount of skepticism and distrust.

Because of the longstanding history of supplier/customer relationships, adoption of different approaches to performance measurement will not come quickly or easily. And certainly, new approaches to performance measurement will not come without the full-fledged support of top management in the partner companies. These leaders must visibly support the new performance initiatives and drive a collaborative culture down through the organization to all levels of operating management. One proven way to overcome mistrust is by creating inter-organizational teams to develop and implement supply chain-wide performance measures. Top managers should set the example by working on collaborative teams with senior executives from partnering organizations.

2.Lack of Understanding

Most managers understand and are comfortable with performance measures that relate to their own company or functional area. Measures that relate to multi-organizational performance, however, fall outside of that comfort zone and invite some negative consequences. For one thing, the measures may never be embraced and, therefore, never be implemented. And even if the measures are accepted, the lack of understanding will make successful implementation virtually impossible.

This hurdle can be overcome with a combination of approaches that begin with top management. If an organization's leaders visibly support the new measurement system and back it up with intensive training, middle management will be more likely to buy in. Further, if the majority of supply chain members buy into the new approach, they can then share lessons learned and effective training techniques among the companies. Inter-organizational teams can participate in joint training programs to build a mutual understanding of how different functional areas fit within the larger supply chain scheme. The training should also emphasize how collaborative behavior produces optimal results for the entire supply chain.

3.Lack of Control

Most managers prefer to be evaluated on measures that are completely under their control. Understandably, they don't want to assume responsibility for results over which they have little or no control. Because many supply chain outcomes depend on inter-organizational efforts, it's easy for supply chain managers to feel that they're at the mercy of someone else.

The challenge from a performance-evaluation perspective is to link individual rewards and recognition processes closely with supply chain-wide performance measures. Without such linkage, individuals will be pursuing conflicting goals across the entire supply chain. Though it may be necessary to combat resistance to chain-spanning measures by retaining certain controllable performance measures, such measures cannot conflict with the broader supply chain objectives.

This is difficult to achieve because, in many instances, chain-spanning measures may require an employee to sub-optimize his or her function's performance to benefit the entire supply chain. For example, an individual planning transportation loads to retail customers would normally seek to ship in truckload quantities to reduce overall transportation costs. However, in the supply chain context, it may be more effective to ship small loads more frequently in order to minimize expensive retail inventories and maximize product availability. If a supply chain goal of high in-stock availability is to be achieved, then the individual reward system for the transportation load planner must recognize that low-cost transportation is not the goal. Furthermore, this individual cannot be penalized for relatively high transportation charges.

4.Different Goals and Objectives

Although it would be nice to think that all supply chain participants are marching to the same drummer, that's typically not the case. The goals of the partners may differ significantly—not for any sinister reason but simply because of the different competitive situations, financial circumstances, and environments they face. If the supply chain goals are not universally accepted, then the chances of agreeing on specific performance measures or on levels of performance are slim indeed.

The best approach to rationalizing conflicting goals and objectives is to involve as many supply chain members as possible in a process of open dialogue. Consensus building is mandatory. If the members cannot agree on goals and performance levels, there is little hope that the supply chain will ever compete effectively. Forming joint teams of individuals from different disciplines and management levels will facilitate the process of establishing goals and resolving conflicts. In the end, top managers from the partner companies may have to sit down together and resolve differences in philosophy, culture, and strategy.

5.Information Systems

Most corporate information systems (IS) are incapable of gathering nontraditional information relating to supply chain performance. Therefore, every organization needs a champion from the IS department who is committed to working with functional managers to create inter-organizational performance measurement and reporting capabilities. In fact, it may be necessary to bring together several IS managers from among the supply chain members to develop common measures and formats. It is encouraging to note that ERP (enterprise resource planning) systems and Web-based solutions now exist to facilitate the task of data gathering and sharing.

6.Lack of Standardized Performance Measures

A supply chain-wide performance-measurement system works only when all of the members have agreed on the proper format, structure, and measurement approach. This may be a significant hurdle to overcome because of the differences between companies with regard to information system sophistication, cost of gathering information, and philosophies on how to measure certain outcomes. Standardized measures can only be created through collaboration—a process that could take several months of negotiation and trial and error. Once the supply chain partners have agreed on the standardized measures, they face the next challenge of developing the actual measurement systems. New measures will require buy-in from top management, programmers, systems personnel, and other affected functional areas.

7.Difficulty in Linking Measures to Customer Value

It is one thing to create a supply chain-spanning performance-measurement system; it is quite another to ensure that the measures used relate to end customer value. Each proposed measure must be carefully evaluated to determine if there's a direct link to the value delivered to the final customer. If the link cannot be established, it may be necessary to discard the measure or to find another measure that does have the required linkage. This is a challenging hurdle to overcome. It requires a substantial investment of people, time, and dollars.

8.Deciding Where to Begin

Deciding where to begin in developing a supply chain-wide performance-measurement system is one of the most significant hurdles to be overcome. In work the authors did with Cooper Tire and Rubber Company, we suggested that the company begin by carefully analyzing key internal logistics performance metrics. Once Cooper agreed upon the appropriate internal metrics, it could then start the process of creating a full-scale internal performance-measurement system that spans functional boundaries. Upon completion of that process, Cooper could begin evaluating performance measures that may be appropriate across the supply chain. Proceeding in this manner is beneficial because many of the internal measures will serve as guidelines for the integrative, chain-spanning measures.

These implementation hurdles are daunting but not insurmountable. If your company and supply chain partners are undaunted by the quest to create a supply chain balanced scorecard, consider the following four pieces of advice.

First, ensure that each supply chain partner is managing its portion of the supply chain in a cross-functional manner. If a company's own procurement, production, marketing, and logistics departments haven't learned to work cooperatively to satisfy customers, how can they work collaboratively to accomplish this across the supply chain? Most companies find that creating a truly cross-functional organization is a major, time-consuming first step. And it's an effort that will utterly fail without the full support of senior management.

The notion of involving senior management leads to our second piece of advice: Don't bother investing the time and political capital in the balanced scorecard initiative if senior management does not wholeheartedly support it. The acid test is management's willingness to replace or supplement functional performance and reward measures with cross-functional ones. If they don't clearly demonstrate this commitment up front, it's futile to continue. Functional management practices will continue to dictate how the organization is run.

If leadership does commit to changing the way in which people are evaluated and rewarded, the company is ready to consider our third piece of advice: Begin the process of formulating a supply chain balanced scorecard by thinking small. Select one—or at most two—supply chain partners to work with initially. Involving more partners than this in the initial stages introduces too much complexity too quickly. The partners should form an inter-organizational team that has cross-functional representation from within each company. One of the group's early tasks should be to define the strategic objectives the supply chain will strive to achieve.

Our fourth piece of related advice is to think small in terms of the number of measures initially adopted. Once the inter-organizational team has reached consensus on strategic objectives, it should identify just two or three measures that will drive the type of collaborative behavior that leads to competitive advantage. The team also should verify that the metrics can be reliably measured before attempting to link them to each company's performance evaluation system. Once the supply chain partners have (1) reliably measured two or three metrics, (2) used the chosen measures to reward employee performance, and (3) confirmed that these measures are driving the intended results, additional measures can be incorporated into the scorecard. Once two or three supply chain partners have developed a complete balanced scorecard, they then can invite one or two additional supply chain partners to participate in the process.

One final observation: When you're working to apply the balanced scorecard to supply chain performance measurement, remember that you're operating largely in uncharted waters. Accordingly, you need to consider yourselves as participants in a learning process, rather than as outside observers seeking to passively benefit from the "best practice" experiences of others. Best practices do not yet exist. The good news is that you and your supply chain partners have a golden opportunity to shape them.