The Elusive Benchmark

We all seek the elusive benchmark that we can compare with others and improve our performance. Benchmarks set standards, define what is achievable, and give us the motivation to be best in class. They are an indispensable part of our management tool kit; yet they are difficult to define, difficult to use, and often inappropriate.

At the simplest level, we can benchmark internally and try to improve relative to our own data and performance metrics. We can also benchmark relative to a select group of noncompeting peer organizations where we share information and best practices, or we can benchmark across the industry as a whole. Each, as shown in the accompanying table, has its advantages and disadvantages. If the benchmark group is small and tightly defined, we have the advantage of accurate and relevant metrics, but lose the opportunity to compare across a wide spectrum of potential winners. If the group is broad, we compare ourselves with all comers but run the risk of using values that do not apply to us because of differences in where we work, how we collect the data, and how we define the metrics.

A good example of the problem is undercarriage life, where industry norms are all but meaningless due to the wide range of conditions under which equipment operates. Narrow the benchmarking group to contractors in Florida, and the values become more accurate and more relevant to companies in the peer group.

Another challenge comes from the way we define the metrics, do the calculations, and collect the data. Availability and utilization are good examples. The illustration (see next page) shows one approach where availability is defined as the time a machine is capable of working divided by the period the machine is required to work and where utilization is defined as the hours worked divided by the hours the machine is capable of working during the day. The definitions are far from universally accepted, and it is impossible to develop availability and utilization benchmarks without first agreeing on the way to collect the data and calculate the required values. Another complication stems from the fact that so many things are interrelated. We can, for instance, see that acceptable availability depends on the length of the required shift as this affects the time available to finish repairs and perform off-shift work within a 24-hour cycle.

The solution is not easy. What can we do to improve our ability to define and use bench-marks?

First, think about and define your benchmarking peer group. Know how to benchmark internally before seeking to develop a peer group or trying to use industry benchmarks. If breakdowns currently generate 22 percent of work orders, then you can set a new benchmark of 15 percent and diligently work toward this new goal. You may not know how you are doing relative to the industry, but you know you will be improving. You will, above all, know that the data you use are good and consistent, and that you are comparing apples with apples.

Second, focus on the equipment-management functions. Define benchmarks that measure performance in field maintenance operations, shop and yard operations, and fleet asset management. These are different functions with different measures for success. Field maintenance operations should be measured in terms of your ability to prevent failures; shop and yard operations are measured in terms of the speed and efficiency with which machines are repaired, rebuilt and made ready for their next assignment; and fleet asset management is measured in terms of financial measures such as profitability and return on assets. High-level financial metrics such as contract value per dollar of fleet-replacement value can be based on industry norms, but even these are subject to discussion on how much work each company subcontracts and how each company values its fleet.

Third, stick to simple metrics. Availability and utilization appear to be simple metrics, but it is difficult to agree on exactly how they will be defined and calculated. Down hours per hour worked is simpler and subject to less interpretation. Define metrics that measure inputs such as mechanic hours per hour worked, outputs such as reliability or uptime, and balance such as the ratio of parts cost to labor cost or the ratio of repair parts and labor to total equipment cost. Remember that physical metrics such as fuel consumption per hour are more reliable than cost-based metrics such as fuel cost per hour.

Fourth, get going. Many companies are reluctant to start a benchmarking and continuous-improvement process because they are constantly searching for a perfect set of universally applicable metrics. The shop overhead data study done by Construction Equipment in June 2005 is a great place to learn about benchmarking and to develop your skills.

Benchmarks tell us where we are now, give us an ability to measure improvement, and reward success. Without them, it is difficult to make decisions and find your way through the maze of available data.