Principles of Measuring Performance

Engineer and management consultant Joseph M. Juran said, “If you don’t measure it, you don’t manage it.” It’s a fairly accurate statement. But, another question might be: “If you do measure it, does that help you manage it?” Far too often, experience shows that it does not, for a host of reasons. Some of these include: having too many measures leading to complexity and confusion about what’s important; a lack of focus; measuring the wrong things; not measuring things that are truly important to the business; having measures that are in conflict across functional boundaries; or not displaying the measures prominently or, if displayed, not keeping measures current, resulting in employees considering them unimportant (after all, if you don’t keep the measures current, how important could they be?).

This article does not describe what you should measure, except perhaps in the form of examples to illustrate a point, but rather shares thoughts to consider about key principles to follow in developing your performance measures.Your performance measures must do the following.

1. Expose Weaknesses

Your performance measures must expose your weaknesses so you can improve them. Another way of saying this is: “Do you want to look good, or be good?” Far too often, measures are “gamed,” that is, jiggered to look good, but on further review are really just a sham. The most common of these is planned and scheduled maintenance. You can have excellent schedule compliance – just don’t schedule much work and leave a lot of buffer time for reactive work or idle time/lost productivity. Your schedule compliance will be excellent, but your costs will be high. The objective of planning and scheduling is to get more work done more efficiently, not to simply look good.

Another common one is overall equipment effectiveness (OEE) and its brother, asset utilization. If you want to show high OEE, don’t use your maximum demonstrated sustainable rate or pick one that’s determined over a long time period to discount all the upsets in your system. And, don’t count short stops, start-up and shutdown losses, planned weekend or night maintenance, and so on. Again, the question is, do you want to look good, or be good? If you want to be good, measure all your losses from ideal and then manage them!

Other examples that might be used to expose weaknesses include: unit cost of production (ideal vs. reality); on time, in full delivery (target of 100%); returns/claims (target of 0); injury rate (target of 0); production schedule compliance (target of 100%); maintenance schedule compliance (100%); and any other measures key to your business success. It’s unlikely you’ll ever achieve any of these levels of performance, but the point is not that. It is to identify your losses from ideal, your weaknesses, and then manage them, making business decisions about the value and impact of eliminating any particular loss. However, all these measures must be balanced against one another, which leads to the next principle.

2. Facilitate Collaboration Across Functional Boundaries

The extensive research of Edgar Schein indicates that the process of organizing creates naturally competing groups – shifts, plants, divisions, etc. That is, people and groups are naturally competitive! He goes on to say that as task interdependence increases, teamwork and collaboration become increasingly critical for organizational effectiveness.

Taking his research to the next logical step then, performance measures must facilitate collaboration, not conflict, across functional boundaries. This principle is particularly relevant in functional groups with high task interdependence, for example, between production and maintenance across shifts, between purchasing/stores and maintenance and between marketing and manufacturing. Far too often, examples where performance measures facilitate conflict across these functional boundaries are evident. For example, when production is held accountable for meeting the production plan with quality product, but this comes at the expense of not doing maintenance in a timely way, that is to the detriment of maintenance schedule compliance (unless you just don’t schedule much work) and to the long-term detriment of the business. Or, when maintenance is held accountable for maintenance costs, but does not control those activities (e.g., design, purchasing and operations) that induce most of the defects that result in a maintenance requirement. Also, when maintenance is held accountable for quality and timely repairs, but does not control spare parts stocks. Another example is when purchasing is held accountable for keeping parts inventory low, but does not control the store room. Or, when marketing and sales do not consider the ability of production to deliver a quality product at a reasonable cost in a timely manner. The list goes on relative to the conflicts that can be created when silos are created and measures within those silos result in conflict, not collaboration.

“Not everything that can be counted counts, and not everything that counts can be counted.”
~ Albert Einstein

So, how do you get collaboration? By having superordinate goals that take priority over group interests, by constantly asking, what’s the right thing to do for the business, and by having cross-functional measures that assure collaboration. For example, to facilitate collaboration between production and maintenance, have one production plan, which includes the maintenance plan, and hold BOTH accountable for maintenance and repair costs, production and maintenance/preventive maintenance (PM) schedule compliance, and on time delivery. To facilitate collaboration between purchasing/stores and maintenance, hold BOTH accountable for inventory turns on parts and stockout rate/service level. To facilitate collaboration across shifts, demand consistency and standard work across all shifts that work in partnership and don’t reward a single shift for performance, but rather reward all shifts for steady performance across all shifts. Other examples could be cited, but this should be sufficient to get your thinking started for using cross-functional measures to facilitate collaboration.

Related to cross-functional measures is the concept that performance measures cascade from the executive suite to the shop floor. As previously noted, they must facilitate collaboration across functional boundaries. And, the same can be said up and down the organization, they must, likewise, facilitate collaboration and be supportive, both upward and downward.

3. Be Highly Visible and Kept Current

Performance measures should be highly visible and their display kept current. Too often, certain measures are kept in someone’s office, operating in a silo mentality and not shared with employees, except to complain or criticize employees for their lack of performance associated with these measures. Alternatively, they are displayed, but are weeks or even months old, suggesting a lack of importance to those who see them daily. If performance measures are not updated in a timely way, they couldn’t be important and employees will behave accordingly.

4. Balance Lagging and Leading Indicators

Performance measures must have the right balance of leading indicators(i.e., the things you do) and lagging indicators (i.e., the results you get). Too often, you see mostly lagging indicators, but this reflects what has already happened and, as the old saying goes, is like looking in the rearview mirror. A far more important issue would be to measure the things you’re doing so you get the right results and then drive those with greater energy.

Here are some examples of leading indicators, but you should select those that you think are most appropriate for you. For operators, they might include:

  • Process conformance/nonconformance;
  • Number of alarms, disabled alarms;
  • Number of spills, loss of containment;
  • Operator care, PM conformance;
  • Equipment downtime, delay times, life;
  • Housekeeping conformance;
  • First pass, first quality yield;
  • Other process specific measures directly influenced.

For maintainers, they might include:

  • Maintenance/PM schedule compliance;
  • Percentage of equipment aligned and balanced;
  • Seal life, number of seals used per month;
  • Bearing life, number of bearings used per month;
  • Lube compliance;
  • Number of leaks per month;
  • Other specific measures directly influenced.

Experience indicates that much more attention needs to be given to the leading indicators, engaging the shop floor in doing the right things. If you do, you’ll get the right business results.

5. Matter

Performance measures must matter, that is, be used for driving behavior. Albert Einstein once said, “Not everything that can be counted counts, and not everything that counts can be counted.” So, you need to do your best to count the things that “count” or matter to the business, while recognizing you won’t be able to count certain things, like culture, except in a very subjective way.

However, you can use specific measures to drive the culture of the organization in a positive manner by engaging the workforce in improvement. People do want to change, if given a compelling reason for change, if there’s something in the change for them and if they participate in developing the changes so they have a sense of purpose and control. Those changes must be measured to reinforce the new behaviors.


Finally, it’s unlikely you’ll be able to get everything presented in this article exactly right. Indeed, these principles inherently lead to imperfections within any functional group. However, if followed, the business, on the whole, will be better for it. You will have made great strides in thinking at a systems level, not a silo level, and your business will benefit from it.