From the Blog

Measure What Matters: Using Metrics to Drive the Right Behaviors

Ruler image
This week’s guest blog entry comes to us from Jeff Sipes, Back2Basics, LLCYou can hear more from Jeff on Lean Metrics, Transformation, Productivity and Continuous Improvement at FABTECH during two Lean Principles Sessions on Friday, Nov. 18.  Learn more here.
Metrics drive behavior.  The question is “does the metric drive the right behavior?”  This is an important question because either the absence of a metric or an incorrect metric can drive unintended behaviors.  Getting metrics right is important to an improvement process. My presentation at FABTECH titled Lean Alignment & Metrics will focus on several techniques to help companies determine what metrics to use and then how to bring the metrics to life.  One technique is to use the “input-transformation-output” model.  Let’s illustrate with a couple of examples:
  • # Materials Shortages:
    • Inputs – open WOs and planned WOs that have material shortages (from manufacturing information system)
    • Transformation – sum of open WOs with shortages plus sum of planned WOs with shortages
    • Outputs – total number of shortages
  • Supplier Satisfaction:
    • Inputs – supplier satisfaction survey/feedback instrument responses
    • Transformation – average of individual respondent scores
    • Outputs – average score and trend line
Now consider the input-transformation-output model in a bit more detail.  It is important to understand where the inputs are coming from.  In the # Materials Shortages example, the data is drawn directly from the information system, preferably with no manual intervention.  The Supplier Satisfaction example is most likely a manual process that is periodically produced (perhaps quarterly).  Is the data readily and consistently available, or do you have to “muscle” the data? The transformation component of the model is where the input data is processed.  In some cases the data may just pass through as raw data with no calculation.  In other cases there will be a specific calculation to develop the output.  Again, the most desirable approach is for any manual intervention to be avoided by having the calculation happen directly from the information system or have the data be directly imported into an Excel file for calculation. The output component of the model provides the result for the metric.  Be sure to understand details like the unit of measure, meaning of trend direction (is up good or bad?), and frequency of reporting. So what do you do with the metric now that it is generated?  Share it, communicate it, and coach/counsel around it.  Make the metric drive behavior the right behavior. I hope to see you at FABTECH in Las Vegas!   [x_author title=”About the Author”]

2024 Platinum Sponsors

2024 Sponsors

Become a Sponsor