Focus and Leverage Part 136

Although I’ve written about performance metrics in past postings, I want to expand upon what I’ve previously written.  Selection of the right performance metrics is critical to the success of any organization, no matter whether they produce products or deliver a service.  There are three key objectives of performance metrics as follows:

1.    First and foremost, performance metrics should stimulate the right behaviors.  Sounds simple enough, but unless the desired behaviors are well thought out, it is very easy to go astray.  For example, if your organization produces products and uses the performance metric operator efficiency, ask yourself what behaviors should you expect to see?  Translated, efficiency deals with minimizing waste and maximizing the capabilities of your human resources.  From a Cost Accounting perspective, it is believed that improving operator efficiency has a direct impact on the profit of a company.  So, if we increase operator efficiency, then we should see a corresponding increase in profits……right?

 

The problem with operator efficiency as a performance metric is that it doesn’t consider the impact on the total system.  The behavior we typically see when using efficiency as a metric is this. Because all of the process steps are encouraged to “run to their maximum capacity” the total system is flooded with excess work-in-process inventory which extends cycle times which negatively impacts on-time delivery.  So do you think efficiency stimulates the right behavior?  The answer is, no it doesn’t.  But having said that, what if we only measured the efficiency of the system constraint?  What would happen then?  We would indeed maximize the throughput of the process because throughput is controlled by the system constraint.  What about the non-constraint process steps?  Because we don’t want them to outpace the constraint, they must effectively “slow-down” so as not to fill the system with excess WIP.  Their efficiencies would deteriorate, but the profit of the overall system would improve dramatically.  This is directly in contrast to what traditional cost accounting teaches.  That is, as operator efficiency increases, profits rise accordingly.

 

2.    Performance metrics should reinforce and support the overall goals and objectives of the company.  If the goal of any for profit company is to make money now and in the future, then the selection of performance metrics must directly support and enhance this goal.

 

3.    The measures should be able to asses, evaluate and provide feedback as to the status of people departments, products, and the total company. The right behaviors of people and departments are critical to the achievement of the overall goal of the company, but many times the metrics chosen encourage and stimulate the opposite behaviors……just as I demonstrated with operator efficiency above.  The fact is efficiency drives local optimization rather than optimization of the total system.

 

When selecting performance metrics, there must be criteria for selection of the correct ones…..right?  Well, in fact, there are at least six key criteria to consider when selecting effective performance metrics.  Let’s look at these criteria and relate each one to operator efficiency.

1.    The metric must be objective, precisely defined, measurable and quantifiable.  There can be no ambiguity at all with the people or departments being measured.  For example, operator efficiency is objective, well defined, measurable and quantifiable, so it would seem to satisfy this first criteria…..right?

 

2.    The metric must be well within the control of the people or departments being measured.   For example, when considering the metric operator efficiency, it clearly is within the control of the people or department being measured, so it would seem that it does satisfy this criteria…..right?

 

3.    The metric must be translatable to everyone within the organization. That is, each operator, supervisor, manager and engineer must understand how his or her actions impact the metric.  For example, efficiency is definitely translatable to everyone, so again, efficiency passes this litmus test….right?

 

4.    The metric must exist as a hierarchy so that every level of the organization knows precisely how their work is tied to and supports the goal and critical success factors of the company. For example, if one of the critical success factors was “maximum throughput” and efficiency was selected as one of the lower level metrics, what would happen?  That is, if it takes 5 minutes to process a part in an individual work station, then the maximum amount of time for the part to be finished should be no longer than 5 minutes.   And if you could somehow produce a part in 4 minutes, the efficiency of that work station could be above 100%.  That would be great….right?  But if the higher level metric was maximum system throughput, would running each individual work station maximize system throughput or would it “clog” the system with excess WIP and cause less than optimum throughput?  For this reason, efficiency is not a good metric because it doesn’t have a positive effect on the system.

 

5.    The metric should be challenging, yet attainable.  Suppose efficiency was selected as a metric.  Is maximum efficiency challenging and attainable?  The answer is no, if you want to optimize system throughput.  Because of step 3 in Goldratt’s 5 focusing steps (i.e. subordinate everything to the system constraint), non-constraints can never be permitted to run as fast as they can or they will choke the system with excessive WIP.  The excessive WIP extends cycle times, ties up cash unnecessarily, and all of the other reasons already mentioned.

6.    The metric should lend themselves to trend and statistical analysis and, as such, should not be “yes or no” in terms of compliance.  We could definitely trend and perform statistical analyses on efficiency.

The point of this posting is to demonstrate why it is so important to select the “right” performance metrics.  The success of organizations is tied directly to the selection of metrics that drive optimal behaviors.  As Goldratt himself said, “Show me how you measure me and I’ll show you how I’ll behave.”  Take a look at the metrics your company is using and see if they pass the metrics litmus test.  If they don’t, then you have a problem.

Bob Sproull