Keeping It Real

I was working with a client to create their business’s critical performance measures the other day, and one of the managers said, “We should set the target higher than we want it—you know, aim high to hit low.” This set off one of my rants, and I thought I would share it with you.

When I worked as an engineer, we would get requests from customers for tighter-than-standard specification limits. I’d do a little control and capability analysis to see if our process was capable of meeting the requested limits. If so, then we would approve the change. If not, I could estimate the amount that would go out of spec over time, and we could calculate the increased cost of scrap due to setting the tighter limits. In some cases, we would go ahead with the change, knowing that it would cost us more money, either because the scrap cost was small or because there was some other reason to accept the loss. (Knowing what I know now about the Taguchi Loss Function, we probably lost money on most of these changes due to the increased nonscrap losses.)

One day I called a customer to say that we wouldn’t be able to meet their tighter limits. “I’m sorry,” I said, “but our process just is not capable of it. Actually, I don’t think there is a process out there that is capable of it.”

“Oh, we know,” he said, “those specs are a lot tighter than we really need. Take the order anyway and just call us if you end up going out of spec and I’ll OK for it to be sent on deviation. We just give those specs to our suppliers so that they try harder. That way, we get stuff that might be out of the spec on the purchase order, but inside of our internal spec, which is what we really need.”

I’m a pretty mellow guy usually, but I just about went ballistic at this point (this was an aerospace application).

While I know that my readers would never do this to a supplier, I bet a lot of you have run into this type of thinking from your customers. Or maybe from your Six Sigma Champions?

So what’s the problem?

Well, first off, I don’t know what his real specs are. So if I keep making stuff out of spec and he keeps telling me to ship it anyway, I’ll eventually conclude that I can cut out the middleman and just not call him when it’s out of spec. Then he might get stuff that really won’t work, but not be warned about it. This teaches contempt for specifications, so congratulations! You’ve just trained someone to ignore them.

It gets worse when a customer gives you a bogus set of specifications and requires a capability metric, say a C_pk ≥ 1. I even had a customer decrease the width of their spec every year while at the same time requiring an ever higher C_pk with the idea that it would “force” our process to get better.

This gets to the heart of the problem, I think. Some people have faith that all they have to do is specify limits and the process will magically get better or that the supplier will try harder and produce better output. This fundamental misunderstanding of how processes improve might explain a lot about some businesses and those who manage them—places where the boss cusses and bangs his or her shoe on the table.

Process output is a function of many things, but not of specifications. Specification limits should be based on what the downstream processes (including the consumer) can tolerate, and better yet on an understanding of the Taguchi Loss Function. The output from a process is what it is, and if your CEO gets into an argument with the process as to what it should be able to produce, the process always wins.

I’m not saying that it’s impossible to make a process better. Identifying and eliminating sources of variability, achieving control, running an experiment to reduce variation, and implementing the findings—these will improve the output of a process. But setting tight tolerances will have no effect but to increase the cost of the process, which is eventually paid for by the very people thinking to benefit from changing the specs.

Similarly, when trying to identify targets for those business-level metrics that monitor the health of the organization, we could choose to set targets well above what we want in order to end up with it. This type of “stretch goal” ought to inspire people, right?

If you do that, everyone knows that it’s a bogus target, and they lose respect for management for setting it. Then the employees have to figure out what the “real” target is, and you can bet that they all have a different idea. And so we are back in the bad old place where we lack organizational alignment, even on something as basic as to how to measure the performance of the organization. Worse, some of the new employees might actually believe management and get stressed trying to achieve the impossible, until they learn the lesson everyone else has—management lies and doesn’t trust us to do the job. You can either bust your butt failing to achieve their unrealistic goals or you can kick back like the rest of us and meet their underwhelming expectations.

Or, you could take a look at the data, decompose the high-level measures into their components, and figure out an achievable target based on where you’ve been in the past and what you can implement to make things better. Then, put together a plan that will achieve those objectives and deploy it throughout your business.

How do people react to that? Do they lack inspiration in the absence of a lofty (impossible) goal? In my experience, they love it. Having a clear, realistic target lets everyone know what needs to change and by how much. Having a plan to achieve it empowers people to…umm…achieve it. I know it sounds weird, but a series of achievable goals tends to get done, while a series of unachievable ones tends to make people tune out and ignore the babble coming from the managers.

The phrase I like about this is “realistic limits strictly enforced, not strict limits loosely enforced.” If you’re given a silly objective and told it is OK if you don’t achieve it, how can you judge what you really do need to achieve? If you are given a realistic objective and you achieve it, congratulations! If it’s realistic and you can’t, well I sure hope you figured that out soon and quickly identified the roadblocks for your boss or Champion to eliminate, because with responsibility comes accountability.

So why am I telling you this? Two reasons: today and tomorrow. Today, I want you to recognize if you’re being put in the position of having unrealistic goals, whether as a Black Belt working on a project or as the manager of an area. Tomorrow, when you’re identifying goals for the people that report to you, I would really feel like I had made a difference if you thought about this topic and kept it real.

About the author
Steven Ouellette is the founder of six-sigma-online.com, president of The ROI Alliance LLC, an instructor at the University of Colorado Engineering Management Program and director of the Center for Statistical Solutions. He has been in process design and improvement since 1992 as an engineer and later as a consultant working in many different industries. He has a Master Black Belt certification, a master’s degree in engineering management, and a bachelor’s degree in metallurgical and material science. He also acts as a board member on Orion Registrar’s committee to safeguard impartiality.

About The Author