Many, if not most, agreements between sellers of goods and logistics service providers (LSPs) contain a list of metrics that are used to evaluate various aspects of supply chain or provider performance. In theory, these metrics should provide clarity to both parties about what success in their relationship looks like. In reality, they frequently breed confusion and conflict. How does this wellintentioned undertaking go so wrong so often? In our experience, Verlyn Suderman it usually has something to do with the parties' failure to address thoughtfully and prospectively the questions of purpose and priority, definitions and methodology, and consequences. In order to avoid these kinds of problems, the contracting parties may want to ask themselves and each other the following questions when developing and documenting their understandings about performance measurement.

1. What do we want from our supply chain?

The effectiveness of a supply chain is determined by how well it enables the brand's success. A supply chain should not exist in a functional vacuum, disconnected in its purpose from the business's core strategic initiatives and objectives. Rather, it should be as closely aligned as possible to the differentiating strategies that the business is pursuing. Ideally, the supply chain function leaders will have been involved in these directionsetting conversations and, as a result, will be able to clearly communicate their vision for organizationally contextualized supply chain excellence. However, even if that is not the case, a best-practice SLA/KPI development process still has to start with those highestlevel questions. Moreover, while those questions ultimately need to be answered by the seller/shipper, that is not to say the LSP should not be involved in the process of identifying firstorder success criteria, especially where the LSP has a substantial scope of responsibility and the ability to derive insights from its activities.

The choice of metrics to be emphasized in any particular relationship depends in general terms on the nature of the operation, the nature of the customer's business, and the customerspecific goals and strategic initiatives. Supply chain success may look very different for an established player in a mature, declining market than it will for a start-up in a fast-growing industry segment. At a more day-to-day level, prioritized metrics may be different for a plantadjacent facility, where avoiding a production line shutdown is paramount, than for an overflow or buffer facility, where storage density may be the main driver of mutual benefit, or a food product mixing center, where timeliness and accuracy in filling retailer orders is the primary focus.

A good provider wants to know what it will take to satisfy its customer, and the customer has every reason to communicate clearly to the provider the handful of expectations and priorities that will make or break the relationship. If the parties apply their best thinking to identify the selective metrics and targets that will drive the service levels desired by the customer and the compensation desired by the provider, they should never experience what has been called a "watermelon scorecard," where the metrics are green but the customer's overall sense of the provider's performance is negative and red.

2. Who are we measuring and with what consequence?

The consequence for any particular performance level will usually depend on (1) how the performance compares to some relevant, objective standard and (2) the degree to which the LSP is responsible for the activity being measured. If the required performance level relates to an activity entirely within the LSP's scope of work and is modest enough that failing to achieve it would be generally seen as unacceptable throughout the industry, then contract default may be the appropriate consequence. If, on the other hand, the customer wants or needs a measure that spans more than one provider's activities (e.g., "Must Arrive By Date" where there are different providers for warehouse and transportation services), accountability in the form of defaultlevel consequences is much less reasonable.

Sophisticated performance measurement regimes often include two or more categories of metrics, each with different consequences. Such a contract might include the following:

  • Core metrics (the 6-10 that are the most important to the business and within the LSP's control), such as:
  1. Breach-level thresholds that represent a fairly egregious failure by the LSP to perform competently and may entitle the customer to terminate and/or recover its damages
  2. Penalty-level thresholds that represent significantly less-than-expected performance and result in some predetermined financial penalty to the LSP but not a breach consequence
  3. Goal-level thresholds that represent expected and satisfactory performance and do not create any contractual consequence
  4. Incentive-level thresholds that represent substantial overachievement and superior performance and entitle the LSP to an additional payment
  • Non-core metrics, such as:
  1. Other metrics that are either less critical or less suitable for accountability purposes, but still have value for management and informational purposes
  2. No contractual consequence except that failure to report may constitute breach

As in the previous section, the goal here is to motivate the LSP to pursue true supply chain optimization, as defined by the customer, by aligning its risks and rewards with the actions desired by the customer.

3. What should our measurement methodology be?

This is the nuts-and-bolts question, and it is often given insufficient attention. Even if the parties intelligently select the metrics that are perfectly tailored to the needs of the business and establish a set of consequences that closely aligns their interests, the system can still fail due to imprecise, flawed or misunderstood methodology. Ambiguous or sloppy definitions or methodology may be the friend of one party or the other in a given situation, but they are always the enemy of a healthy performance management system.

Questions to be answered related to measurement methodology include:

  • What are the business rules and exceptions?
  • Whose system provides the calculation inputs? Who does the calculations and reporting?
  • What is the measurement period? Is there a cure period?
  • To what extent do the required performance levels apply during start-up?
  • Under what circumstances are required performance levels subject to adjustment?
  • Are there any guiding principles that govern the administration of the performance management regime?

No system will be perfect, but putting in the effort up front to address as many potentially problematic situations as possible is always useful. It may be unrealistic, for example, to expect to define a weather exception to an ontime delivery metric in completely unambiguous terms, but both parties are better off, at least prospectively and relationally, if their definitional work eliminates as much ambiguity as reasonably possible.

In summary, supply chain performance metrics, whether they are called SLAs, KPIs or some other name, can be an invaluable tool in harmonizing the resources and efforts of a seller/shipper and its LSP to create competitive advantage. However, they will fail to deliver on this promise if (1) they are not developed with sufficient knowledge of, or connection to, the seller/shipper's strategic objectives, (2) they do not motivate the right LSP behavior, or (3) they are not defined in such a way that uncertainty in application and opportunity for manipulation are minimized.