In the past few months we have reached out to our clients asking them to share their unbiased experience of launching and using Six Sigma at their firms.  Here is a wonderful learning lesson shared by a Master Black Belt at a large lending firm.

If You Can’t Measure It, Don’t Bother Trying to Improve It 

To a person who has been involved in Six Sigma for some time, this is likely to be a mantra; a sort of Golden Rule not to be violated under any conditions.  To others, some might think it is a statement born out of common sense, while yet others are surely saying that something that cannot be measured is precisely something that should be improved to the point that it can be measured.  In a sense, all of you are correct, but there are pitfalls to navigate.

Starting with the basics, it’s essential to understand that the single fundamental assumption in Six Sigma methodology is that the inputs and outputs of any process as a whole and even the individual steps of a process can be precisely observed, measured, and reported.  If any part if missing, then you don’t have a Six Sigma project in the true sense, because you cannot state the problem in terms of a measurable defect, you cannot calculate the potential improvement, you cannot state the expected outcome the project in a measurable way, and consequently, you will never know if the project was ever truly successful, or even if the project has reached a conclusion of some kind, successful or otherwise.

I remember a project I supervised while working for a commercial lender.  We had observed for some time that there was tremendous variation in the cycle times for loan origination.  This was creating inefficiencies in a number of areas with direct impact to the income statement, including employment costs, treasury cost, and cash management to name a few.  As we discussed the best way to tackle the problem, the greatest challenge facing the Black Belts was the fact that all employees were salaried employees, and while the company did have macro data regarding total cycle time, we did not have any data at all about the time spent on each step in each stage of the process.  If it was you in this situation, what would you do to overcome this problem? 

Without some way to gather data about how much time each person touching the process spent at each step in each stage of the larger process, it was impossible for us to undertake a DMAIC project to reduce variation in the cycle times to originate our loan products.  (Incidentally, we wanted to reduce that cycle time overall as well—a Lean concept, but I want to keep thing simple for this example.)  Consequently, we lost a full year of time as we had to develop a data gathering mechanism and procedures to accumulate the missing data to support the subsequent development of the DMAIC project.

Gathering data about a process is almost always a tedious and expensive proposition for a company.  Technology can overcome some of these negative points, but only if the systems supporting the process in question were designed to gather, store and report data about the process.  If not, you are facing an uphill battle to get the data you will need.  You will encounter ardent resistance from those employees impacted by the changes in process required to gather data, as it typically adds steps to the process and increases the complexity and need for attention to detail in their jobs.  You will also encounter resistance from managers who either do not understand the need for the data or are too short-sighted to care.  These are the most dangerous situations, because you will typically experience pressure to make improvements in these areas where you have no data, and will feel you need to do something to show results for your group.  Don’t give in to this temptation!  It is the greatest mistake that you can make. 

So the most important question you can ask yourself or your group if you are leading a project selection discussion is whether the necessary data is available to support a DMAIC project.  If not, can the data be obtained within an acceptable budget?  If the answer is still no, then you cannot proceed, no matter how desperately you would like to improve a process.  Your choices are limited to devising a data-gathering mechanism with acceptable costs, or designing an entirely new process or supporting technology that will meet your data needs.  Incidentally, this is why it is so essential for the Six Sigma team to be involved in the design and implementation of any new process or technology in the company.

S. Shahbazi ProcessArc, Inc. - Financial Services Six Sigma