Significance of Measurement
Over the last several months I have spent a significant amount of time working with various businesses to better define their dilemma. Dilemma is a challenge, goal or objective that we are working to achieve, such as additional impact to the bottom line. Our support has been provided from both a manufacturing and a transactional point of view. Even though these environments seem to be uniquely different, at the end of the day each environment is defined by a series of processes which both have plenty of that. After completing several client dilemma discussions, the challenge that I found in developing understanding the actual dilemma is down to several reasons. Some reasons are unique to each particular business, others are more common across most businesses. The single most common theme we found in understanding dilemma centered on the metrics that they are looking at. People get so consumed in the metric or measurement that they forget the uncertainty, risk or input they are reviewing.
Now, you may think I am going to say all metrics are bad and Lean companies should be able to visualize how their company is running. Well, in fact this is not actually our viewpoint. Based on my Six Sigma background, I know very well that metrics are extremely powerful, when used correctly. The challenge is actually utilizing the metrics correctly is the problem faced by many of the companies we have been working with. The incorrect use comes down to one of three reasons:
The first is the most common reason for the incorrect use of metrics. It is as simple as the amount of measurements companies require to be made. I recently come across an article in LinkedIn by Bernard Marr that discusses 75 KPIs every Manager Needs to Know. While I agree with the premise of the authors discussion that focuses on understanding all metrics and selecting a critical few that will really make a difference. All too often we find it is difficult to select the right measures because we first don’t understand the actual dilemma that is facing the business, operation, process, company, etc. At this point we haven’t defined the specific uncertainty that we are trying to reduce. We then start to select what we think makes sense and measuring it one way. This ties directly back into the dilemma, we get so caught up in the metrics that it is difficult to see the actual dilemma we are trying to solve. The proverbial cannot see the forest for the trees. The starting point should be to determine what are the critical success measures? Every business is different so I cannot say 5, 10 or even 20. However, I do know it is not 75.
The second reason emphasizes more on how often each measurement conflicts with another i.e. one measure being driven positively has a direct negative impact on another measure. A common example we continually come across in our work within Supply Chain networks is that Purchasing agree a low piece part price based on significantly large minimum order quantities that then negatively impact inventory reduction and optimization. This one can be tricky to define, however Hoshin Kanri provides us a great tool to help visualize all of your metrics and how each tie to the strategies of the business and the potential conflicts that might arise from these measures. It is a wonderful tool for outlining your dilemma as it brings everything that the business is focused on all in one matrix.
The third and final reason for the incorrect use of metrics is based on how the numbers are being measured i.e. looking at average. Using only one measurement to understand how you are doing versus the dilemma is a mistake. I can have a really great average and terrible performance due to the variability around my performance. In Production we can have an average cycle time, let’s say 5 days with a target of 5. AWESOME! We are doing great. When you start to dig deeper into the cycle time data you then see that it actually takes between 1 and 10 days with the majority being at both ends of the spectrum, obviously not a DFT environment. In this case the average is calculated to be 5 days showing we are winning and yet we can’t understand why the Sales Department tell us we are not consistent. However, when we calculate the Standard Deviation for the collected data, we find it is also 5; a significant problem. This example clearly shows how measuring the wrong way will actually make us think we are performing well when actually our process is broken.
Each situation explained above is a great use of measurement, where each provides us information about our business. I think Douglas Hubbard the author of “How to Measure Anything” put it best when he defined measurement as a “quantitatively expressed reduction of uncertainty based on one or more observations”. We capture data points to tell us how we are reducing uncertainty. When reducing uncertainty you must always remember, why. I frequently see teams rushing out to measure something before they have ever really understood why they are measuring. Simply asking the question “why are we measuring” seems to be an afterthought to the fact that something is being measured. We consistently never take the time to fully explain or understand our dilemma before something is measured. Metrics for the sake of measuring if you will. This is where other tools such as PDCA, Kaizen and DMAIC are all found to be very useful. Each of these different approaches force us to first understand the why, the dilemma, the reason for measuring before we actually try and create a new metric to reduce uncertainty. It’s the start of 2016, is it time to ask yourself, should we evaluate all of our measures and see what makes sense, what dilemma are we trying to solve in 2016?
Hubbard, Douglas W. (2014). How to Measure Anything: Finding the Value of Intangibles in Business (p. 31). Wiley.
Marr, B. (2013) 75 KPIs Every Manager Needs to Know. LinkedIn.com Retrieved on Jan 23rd, 2015 from https://www.linkedin.com/pulse/20130905053105-64875646-the-75-kpis-every-manager-needs-to-know