Motivating Supply Chain Behavior: The Right Incentives Can Make All the Difference

Incentives do indeed drive operational performance. But how do you determine the right incentives and what measures do you use to make sure they're working?

By Shoshanah A. Cohen, Susan Kulp, and Taylor Randall -- Supply Chain Management Review, 5/1/2007

• The Problem with Incentives
• The Cause of Clashing Metrics
• Organizations and Decision Control
• What the Study Uncovered
• Lessons Learned

It has never been easy to get incentive compensation right. The subject brings up all sorts of arguments over fairness and measurement systems, and can become very complex very quickly.

Formulating and managing incentives holistically throughout the supply chain is particularly challenging. The most pervasive problem is that in many companies, the supply chain is organized as a set of functions or departments such as receiving, production, and logistics. Each function is likely to use metrics that optimize performance inside its box on the organization chart. But when managers' bonuses are based on criteria that make sense only within their own departments, the resulting behavior may conflict with the interests of the company as a whole. Even organizations that are more tightly integrated may find it difficult to select metrics that drive behaviors that are good for the whole company.

That said, incentive mechanisms now deserve renewed scrutiny. The reason is simple: the economic stakes are getting higher as the supply chain itself becomes a more powerful strategic lever. As supply chain management moves higher on the executive management agenda, leading companies are paying closer attention to goal alignment, so that the supply chain can make the business as effective as possible.

Although opinions on the effectiveness of different incentives vary, there is no question that well-designed incentives—those sensitive to the employee's effort, measured relatively precisely, and consistent with firm objectives—do indeed influence the desired behavior. Academic studies confirm that more difficult tasks require stronger incentives, and that managers do need larger incentives to reach tougher goals. The incentives can take many forms: The most common type is the management bonus, which is typically awarded according to a regular schedule and upon achievement of specific objectives.

But big questions remain. Is it simply a matter of establishing the right incentives, with good performance sure to follow? A direct connection between effective incentives and higher performance has always been assumed. Discussions about the difficulty of selecting appropriate metrics led to the research described in this article. Further conversations resulted in more questions: What effect do incentives really have on the achievement of performance objectives? Does it matter where in the organization the incentives are applied? Does it matter how the organization is structured? Can incentives drive a function to perform better even if it doesn't have direct control over the associated operations—as in the case of, say, the manufacturing department's influence over delivery performance?

To answer these questions from the perspectives of both practitioners and academics, we developed the Survey on Information, Incentives, and Performance Management within Supply Chains, a comprehensive study of the use of incentives and the associated business performance. The survey was designed to assess how organizations use supply chain metrics to communicate information and motivate desired behavior. We used the results to calculate the correlation between the various uses of supply chain metrics and operational performance. Completed in late 2006, the survey included mid- to senior-level supply chain executives at nearly 50 organizational divisions representing a broad range of industries. Participation was limited to companies that primarily employ a make-to-stock manufacturing strategy.

While some of the conclusions were consistent with our expectations, others were more surprising. All can be used to increase the effectiveness of supply chain performance management programs.

The Problem with Incentives

An incentive, the proverbial “carrot,” is not worth trying if there is no way to measure its effectiveness. It ought to be easy to do so: simply measure whether the process that the incentive was designed to improve has actually gotten better.

While this sounds straightforward, it requires that companies have effective performance measurement systems in place—and that they monitor them regularly. While most people agree with the adage “If you can't measure it, you can't fix it,” we find that few metrics programs actually provide a clear picture of overall performance, pinpoint the root of performance problems, or identify improvement opportunities. The reason is simple: It's really tough to establish a performance measurement program that is truly robust and useful! Just reaching a consensus on what to measure, how often and with what metrics can be a major effort. And getting management to agree on the fundamental purpose of a metrics program can be the hardest job of all.

Consider a typical example. One of PRTM's clients had serious problems with its order management process. Specifically, customers were complaining that orders took more than 20 days to process, but no one really knew the true cycle time. Why not? Because managers resisted putting formal measurements in place. They were concerned that monitoring the specific processes would highlight each department's problems — and be used by the executive team punitively against the respective managers. It took a concerted effort to help the managers to see the real purpose behind a metrics program: to get a clear picture of actual activities with the objective of fixing the order management problem.

In worst cases, supply chain metrics can serve as a competitive weapon between organizational units. This happens when one department uses metrics to highlight the failings of another (usually to distract from its own poor performance). More commonly, companies find that their metrics reflect conflicting objectives between the various functions. Consider the well-known tradeoff between inventory investment and customer service levels. If inventory targets are lax, achieving superior delivery performance is easy. By contrast, a company may aim to keep its inventory low and establish metrics that identify and remedy inventory levels that exceed targets, thereby making it harder to meet demand.

Just one example: A leading maker of computer peripherals developed a strategy focused on lowest-cost production, constant innovation, and a make-to-stock approach for fast fulfillment, with products ready to ship within two to three days of customer order receipt. Supply chain metrics included product cost, delivery performance, and fill rate. The company set up plants in locations with low labor rates and developed long-term contracts with ocean carriers to ship to distribution centers (DCs). Although most of the factories were in Asia, most customers were in North America and Europe, so products took up to five weeks to reach the DCs. This made it difficult, if not impossible, to achieve the objective of fast order fulfillment. Accurate demand forecasting was already extremely difficult in the volatile peripherals market. And constant product introductions and phase-outs complicated things further.

One consequence of inaccurate forecasting was that the company frequently had to expedite order fulfillment, using airfreight to maintain customer service levels and nearly tripling the company's transportation costs as a result. The product managers didn't see a problem. Since the transportation and rework costs appeared to them as period expenses charged to the supply chain organization, they saw no changes to their product cost metrics.

Of course, the total cost of managing the supply chain increased significantly because of these unplanned expenses. To address this problem, the management team began measuring total supply chain management costs each quarter and looked at how the supply chain strategy affected costs related to order management, materials acquisition, inventory, and planning. The managers also worked closely with each product group to communicate the importance of seeing the total cost of supply chain management. Product costs were still measured regularly, but the entire company was asked to focus on the new metric. As a result, product managers began to see the huge expense associated with expediting shipments, the bulk of which stemmed from planning process issues.

The Cause of Clashing Metrics

Our peripherals company is typical in that many of its conflicting objectives were the direct results of its failure to forecast demand accurately. Companies lacking accurate forecasts are forced to rely on inventory and expediting to meet customer requirements. And because the associated costs are difficult to tie to individual products, business units may not consider them at all when making decisions about how much inventory to hold or whether to use expensive shipping modes. Accurate forecasts drive better business performance because they give the supply chain organization a heads-up on what will be needed and when. Companies with better forecast accuracy show better performance across several important performance indicators, including on-time delivery to commit, fill rate, earnings before interest and taxation (EBIT), and finished-goods inventory levels.

So why not provide a strong incentive to motivate better forecasting? The simple truth is that predicting the future is extremely difficult. While numerous tools are available to assist with analysis of historical data and market trends, developing a forecast that goes beyond a simple extrapolation of recent trends is both an art and a science. Does that mean that even a powerful incentive is not likely to guarantee success in forecasting?

This question was one of the original catalysts for our research. In company after company, we had seen one function held responsible for developing a sales forecast while another was chartered with meeting inventory targets. We also noted that organizations seemed much more likely to measure compliance with, and provide incentives for, meeting inventory targets than forecast accuracy. Our experiences with numerous organizations showed a clear trend: Even when the forecast developed by one functional group was not particularly accurate, the other functional group was expected to meet its inventory objectives. A bad forecast was not a good excuse for having too much excess or obsolete inventory, nor was the group that produced the bad forecast held accountable.

Organizations and Decision Control

Academicians define two major categories for decision-making within organizations:

1. Centralized control: One function or individual is responsible for decisions relating to multiple aspects of business operations. As an example, in the context of supply chain management, the operations or supply chain function may be responsible for forecast development in addition to supply planning and inventory management.
2. Decentralized control: Individuals have clear functional specialties and their responsibilities and decision-making authority are tied to these specialties. In an organization with decentralized control, the sales or marketing function may have responsibility for forecasting, and the supply chain function may be responsible for supply planning and inventory management.

Our survey seeks to determine whether incentives are more effective for improving performance within centralized or decentralized organizations. We classified each company's approach as “centralized” or “decentralized” based on its answer to the question, “Is the same function responsible for both forecast development and achieving the inventory turns target?” A “yes” response indicated centralized decision-making. A “no” implied a decentralized organization. The majority (72 percent) of the survey respondents turned out to be decentralized organizations. (Note that in this discussion the terms centralized and decentralized refer to the structure of supply chain operations. In this context, they are unrelated to the actual organization structure with respect to the existence or locations of corporate divisions.)

What the Study Uncovered

Our analysis of the survey results yielded the following observations:

• Incentives do indeed drive improved operational performance.
• Incentives are more effective for tasks that are more easily controlled.
• Well-developed systems for monitoring and providing information about supply chain performance can be substituted for incentives.
• Companies with higher supply chain complexity are more likely to use incentives.

Here is a closer look at the results:

Multi-Functional Tasks = More Incentives

To gauge the level of incentives deployment, survey respondents were asked to indicate the extent to which incentive compensation for senior functional managers is affected by key performance indicators. These key performance indicators are categorized as either global or local metrics. Global supply chain metrics focus on overall organizational outcomes and are influenced by multiple functions; they include inventory turns, on-time delivery, and total supply chain management costs. Local metrics, associated with more isolated tasks, maximize the outcome of a given function, where individual or departmental effort is more relevant. They include order fulfillment lead time, forecast accuracy, and inventory obsolescence.

Naturally, the supply chain organization is far more likely than other functional groups to have incentives associated with key supply chain metrics. We also found that organizations are more likely to focus on global metrics with respect to incentive compensation plans. Discussions with selected survey participants indicate that this is related to the fact that achieving local metrics targets is a basic expectation of most supply chain positions. Success simply indicates compliance with anticipated competency and bonuses are not typically awarded for meeting fundamental expectations. Global metrics, on the other hand, require collaboration with other functions and the ability to influence the behaviors beyond functional boundaries. Incentive compensation provides motivation to do so.

Most Companies Measure Supply Chain Performance

The organizations in our survey appear to have taken on the challenge of establishing an effective performance management program. In fact, 85 percent of the respondents report that their company uses a supply chain scorecard or other regular means of measuring operational performance. While the contents of the scorecards vary, 89 percent of the respondents measure on-time delivery as a critical operational metric, and 79 percent measure inventory turns.

The scorecards are used for everything from simply informing management about recent performance to providing specific incentives, or even enforcing management policies. Not surprisingly, most companies use operational scorecards primarily as monitoring mechanisms — and the supply chain organization is the most likely area to be monitored. 3Com Corp. is a good case in point. The company created a supply chain scorecard that measures critical aspects of performance as well as more detailed metrics that provide broader visibility into overall supply chain health. Performance metrics include delivery predictability, stock-out percentage, order cycle time, and supply chain costs. These costs can be broken down into material costs, overhead costs and period costs, revealing even more detail.

We hypothesized that the operations/supply chain group is expected to meet delivery and inventory targets, even though it might receive highly inaccurate forecasts from another function. The data confirms this: Even in decentralized organizations, where the operations function is not responsible for forecasting, it is still the most closely monitored. This finding was of special interest to us, given that the majority (73 percent) of centralized companies measure forecast accuracy regularly while only 41 percent of decentralized companies do so.

It is not surprising that scorecards are used primarily to inform and monitor. Many supply chain studies cite information as a critical factor in efficiently managing inventory. Other research has shown that the use of information to inform and monitor is often a catalyst for performance enhancement. Our survey findings are consistent with these conclusions. In essence, information can serve as a substitute for incentives because it alone can motivate managerial effort.

The right set of metrics can tell you how well each supply chain process is performing, highlight where there's room for improvement, and help diagnose problems and decide where to focus improvement efforts. Metrics can also be a powerful management tool by letting people know what is expected of them and allowing you to track progress (or lack thereof) over time.

The knowledge that the act of measurement can improve performance should encourage more companies to monitor their forecast accuracy (currently only half of the survey participants do so). Given the strong correlation between precise forecasting and operational effectiveness, many organizations might improve their performance simply by emphasizing the importance of forecast accuracy.

Higher Complexity = Greater Use of Incentives

While companies typically award performance-based bonuses as part of an individual's performance review, incentives can also be used to encourage the effective execution of coordinated tasks. For example, a supply chain organization might use incentives to coordinate inventory management, customer service levels, forecast accuracy, and order fulfillment lead time. As we saw earlier, failure to align critical supply chain metrics can result in local optimization that is detrimental to overall company performance.

For this reason, the use of performance incentives is particularly important in highly complex organizations facing difficult supply chain coordination challenges, such as those whose supply chains are tightly linked with those of their customers and/or suppliers.

Supply chain executives point to rampant product proliferation as another driver of rising complexity. Ever-expanding product portfolios lead to higher costs, decreased flexibility, and lower quality within both products and processes. The challenges of supply chain management are more difficult for organizations that face:

• High demand uncertainty in which demand fluctuates widely among customers.
• High product complexity.
• Highly aggressive targets for critical supply chain metrics.
• Extensive collaboration with customers and/or suppliers.

Our study found that companies that facing demand uncertainty don't necessarily make more use of incentives. The story is similar for companies that have tough targets for their supply metrics. On the other hand, companies with high product complexity and an extensive collaboration with supply chain partners do tend to use incentives more than other organizations. Our analysis indicates that this is related to the same phenomenon that drives higher use of incentive compensation for global metrics. Cooperation beyond organizational boundaries is required. In other words, it is not enough to be good at your job; you have to help others be good at theirs.

Centralized Control = Less Use of Incentives

There are many ways to coordinate and motivate employee effort within supply chain operations. Senior management may choose to avoid the coordination task altogether and centralize the supply chain function, making one group responsible for most decisions relating to most supply chain management activities. Such a structure, combined with ongoing monitoring of supply chain performance, can serve as a substitute for formal performance incentives. Indeed, our survey shows that organizations with centralized structures and well-developed supply chain monitoring systems tend to use fewer incentives than companies with decentralized structures and less frequent performance monitoring.

The centralized organizations tend to use global incentives such as inventory turns, on-time delivery, and total supply chain management costs to maximize alignment, coordination, and information-sharing. Companies with highly complex products are also more likely to use global incentives since they require greater internal coordination and cooperation to manage their complex supply chains.

Decentralized Control = Better Customer Service

A common academic theory in operations management holds that organizations with centralized controls outperform those with decentralized decision-making. The premise behind this theory is fairly simple. Decentralized decisions are made by numerous individuals, and therefore are less likely to be based on a consistent strategic viewpoint and data analysis. Conversely, in a centralized organization, one person aggregates information from multiple departments and uses this common data pool to make decisions consistent with the overall business strategy.

However, other recent research finds that decentralized organizations outperform centralized organizations under certain conditions. This is particularly true when functional managers make decisions that consider the big picture rather than just focusing on their own department. Thus one way to eliminate functional bias is to provide the decision-maker with incentives that benefit the whole organization-in other words, to align incentive compensation with global metrics.

So which model is better? We find that a decentralized model, coupled with an effective incentives program, boosts overall performance. Our survey shows that companies using this combination do better in such metrics as on-time delivery and forecast accuracy. Fully 55 percent of decentralized companies deliver on time at least 90 percent of the time, whereas only 34 percent of centralized companies do so.

Incentives Do Work

Since we know that incentives should be effective in influencing performance, it's reasonable to expect that companies using incentive compensation for forecast accuracy, inventory turns, and on-time delivery would have better forecast performance, inventory turns, and delivery performance respectively. Survey results confirm these hypotheses, although we find there is a higher correlation between incentives for inventory and delivery performance and operational performance than for incentives related to forecast accuracy. This is likely related to the inherent nature of the forecasting process; provision of incentives is not enough to compensate for the difficulty of predicting customer behaviors and buying patterns.

But we also find that companies aren't any more likely to provide incentive compensation related to forecast accuracy than they are for other operational metrics. In fact, sales-the function most likely to have responsibility for forecasting-is much less likely to have any incentive compensation tied to its forecast accuracy than, say, the operations group.

Lessons Learned

The survey confirms that incentives designed to boost inventory and delivery performance are indeed effective. We also find that providing incentives for forecast accuracy does improve forecast quality, albeit with a weaker correlation. This is particularly evident in companies that use a decentralized organizational model, which have worse forecast accuracy than their centralized peers. These companies can improve the quality of their forecasts by offering incentives to the organization responsible for forecast development.

What does this all mean to the executive responsible for the performance of the supply chain organization? Here are four key takeaways:

1. Have a balanced supply chain scorecard in place and monitor it regularly. We have seen numerous examples of highly detailed and regularly published operational reports that are rarely reviewed or leveraged to improve performance. Our research shows that in many cases, the act of rigorously reviewing operational results can be enough to improve the performance-a non-financial incentive, if you like.
2. Align the responsibility for achieving specific targets with the responsibility for the related operational activities. If, for example, the supply chain organization is not involved with the sales forecasting process, it should not be measured on that forecast's accuracy. Conversely, if the sales organization is responsible for developing the forecast, it should share some responsibility for the resulting inventory investment.
3. Align your use of incentives with your organizational model and your ability to measure performance. If you use a centralized organizational model and have well-developed systems for monitoring your supply chain performance, you may not need additional performance incentives at all. If, on the other hand, your organization has a decentralized structure and you don't rigorously monitor performance, you should absolutely consider implementing performance incentives-making sure that the incentives are aligned with specific supply chain metrics and that you can monitor performance accurately. This is particularly important for incentives associated with forecast accuracy; there's no point in developing an incentive program related to forecast accuracy if you cannot measure it.
4. Apply the incentives in ways that promote the behavior you want. Just as you can organize supply chain tasks in any number of ways, there is no one-size-fits-all approach to incentives. Be thoughtful about when and where incentives are applied, and choose metrics designed to minimize any functional biases. Do not rely on incentives to drive performance within a function that has no control over the metric you intend to affect. Remember that incentives work best when the tasks they are associated with are controllable by the people for whom the incentives are designed.

---

Author Information

Shoshanah Cohen is a director at PRTM Management Consultants. Susan Kulp is an assistant professor at Harvard Business School. Taylor Randall is assistant professor at the David Eccles School of Business, University of Utah.




---

Sources:

Strategic Supply Chain Management: The Five Disciplines for Top Performance, Shoshanah Cohen & Joseph Roussel, McGraw-Hill, 2004 Forecast Accuracy Measurement and Management, as presented by Rick Hoole at the Institute of Business Forecasting Supply Chain Forecasting Conference, February 2006 Strategic Supply Chain Management: The Five Disciplines for Top Performance, Shoshanah Cohen & Joseph Roussel, McGraw-Hill, 2004 Drive Complexity out of Your Supply Chain, Rick Hoole, Harvard Business Review Supply Chain Strategy, 2006 Alonso, R., W. Dessein, N. Matouschek. 2006. When does coordination require centralization? Working paper, Northwestern University.

Talkback

We would love your feedback!

Post a comment

» VIEW ALL TALKBACK THREADS

Related Content

• Topics
• Author

Related Content

 

By This Author

There are no other articles written by this author.

Sponsored Links