"Frustrating" is the word Duke Okes uses to describe his reaction to the question, "What performance measures should I use?" The author of "Performance Metrics: The Levers for Process Improvement" doesn't get frustrated because it is a bad question to ask, but because it should never be the first question you ask, and the answer always is "It depends."
See Also: Operations Management Strategy & Best Practices
It depends, he says, on what you are trying to accomplish. "Ask 'What is the direction I am trying to go in? What is the mission and strategy of my department or operation?'" he says. "What you measure will be determined by the answers to [those questions]."
With "Performance Metrics" (2013, Quality Press), Okes focuses not on high-level metrics, such as revenue or profit -- "most organizations have those" -- but on the lower-level metrics that drive the high-level outcomes.
In addition to the basics, Okes' new book addresses what he says is an often overlooked factor in discussions about performance metrics -- the psychological impact. Quite simply, he says, "they often scare people because they can impact pay, performance and even the sense of self." Given that, Okes suggests organizations not simply impose metrics and demand they be followed, but instead test them. "Don't just expect people to understand metrics and use them effectively."
Be aware as well, he says, of the games people can play with metrics, and put measures in place to prevent it. He cites the example of a purchasing department being asked to perform to a lowest-cost metric. In their drive to meet that measure, buyers procure components that achieve their goals but create chaos down the production line due to quality or delivery issues.
Okes advice: Make sure the metrics you select don't mistakenly promote individual- or department-level performance to the detriment of the overall organization. In the purchasing example, a better system-level metric may have been total cost of ownership.
For many organizations, numbers are what trip them up when it comes to addressing performance metrics. How many metrics should you follow? How frequently do you measure them? Technology has only eased the ability to collect vast quantities of data, whether they are needed or not.
That ease propels Okes to issue a caution. "One of the things we should be paying more attention to is the cost of metrics. Since many systems are computerized, we can measure everything, but there is a cost. There is a cost to write code, a cost to set up the system, a cost to acquire and process and report and review and make decisions," he says. "The potential cost to making a bad decision is huge."