Lean and Quality – a Match Made in Manufacturing Heaven

“Lean isn’t going to help you if your quality is bad. Why just make bad stuff faster?”

I hear this comment, or ones like it, more often than I’d like. It’s a misnomer that for whatever reason has caught on, even though it couldn’t be more wrong. Lean is a lot more than just “better, faster, cheaper.” The fact is that Lean and Quality are close cousins, and the relationship needs to be better understood throughout industry.

A former employer of mine learned this lesson the hard way. The manufacturing of one of their products was done in Puerto Rico, but product testing (which was performed on 100% of the units) was being done in Connecticut, so right away we knew we had a “waste of transportation” problem. To give you some perspective, units were about the size and weight of a window air conditioner, so shipping them was costly, and a bit of a pain!

Units would be assembled per specification with vendor certified parts and then shipped to us for testing. Testing each unit took hours, and they would fail about 50% of the time, so back to Puerto Rico they would go for rebuild and/or repair. Units would then be fixed and shipped back to Connecticut for another round of testing. Again, 50% of those units would fail, and the process repeated itself until a unit passed. For those keeping score: the all-time round-trip record for a unit was seven trips between PR and CT until the unit finally passed.

As all of this was happening, our customer was waiting for their products to be delivered, and we were falling farther and farther behind their schedule. Weekly update phone calls quickly turned into daily ones, which eventually devolved into little more than our customer screaming obscenities at us before hanging up. It wasn’t a fun time – for anyone.

In an effort to gain some breathing room, we tried increasing our throughput at the plant. We increased overtime, hoping that an increase in direct labor would help us produce more, and “catch up” to the schedule. We also increased our raw material inventory to such high levels that we were stuffing it anywhere we could find space. We tossed any semblance of lot control out the window—if we came upon a new lot of materials that seemed to “be good,” we sucked it into production as fast as we could, pushing all others to the side (for the record, this helped us increase first-time test yield in the short term, but the long-term remained right around 50%—go figure!). Truly, we thought we were doing whatever it took to try and please the customer. We knew we had problems, but given the time and resources we had, we made the calls we did thinking that if we could run parallel paths (increase throughput to get ahead while troubleshooting the product), we’d eventually work ourselves through the issue.

In retrospect, we should have done a lot of things differently right from the start. We didn’t follow any kind of formalized problem-solving methodology, we didn’t track test data, and we had poor controls in manufacturing. When units went back to Puerto Rico to get repaired after failing test, there were no instructions or guidance on why the unit had failed or what it had failed for, so the folks there would fix what they thought was the most likely problem component and send it back to Connecticut to see if what they had done worked.

So, considering all of this, would you say that we had a Lean problem, or a quality problem?

It was both, really. Having interdependent groups so far apart in a matrix organization all but guaranteed that miscommunications and misunderstandings would occur. Trips back and forth (for people and for product) were lengthy and expensive. Incomplete data and record keeping made it impossible to properly perform any kind of root cause/corrective action. Increasing raw material levels also caused more problems than it solved. We were tying up more cash, and were literally tripping and knocking over raw materials on the plant floor, undoubtedly causing even more damage. Running oodles of overtime in a futile, non-descript manner meant that folks were getting tired, less productive and morale was suffering. More mistakes than usual were being made that weren’t caught until a unit had made the 1,650-mile trip to Connecticut.

Lean isn’t about speed—it’s about value. Those are two different things that are often mistakenly interchanged. If a customer understands and agrees that a process adds value (and will pay for it), it doesn’t really matter how long the cycle time takes (assuming it’s not preventing you from meeting their needs). Of course, it helps to look for ways to reduce overall cycle time if you can, but more often than not reducing cycle times aren’t anywhere close to being your biggest challenges to adding value. Instead, it’s all of the “other” parts that make up your lead time that need help: the wasted time and effort things spend waiting, being moved, being over processed and overproduced. In our example, we did all of these things, in addition to adding extra inspections and test protocols thinking that those added operations would somehow help us fix our problems. All they really did was slow the process down even more than it already was, and kept us from focusing on other, more important things.

Quality is about value as well, albeit from a slightly different angle. All quality comes at a cost, whether it be good (prevention, appraisal) or poor (internal failure, external failure). Each one of those costs represent different levels of value to the customer, and to the parent organization. Customers won’t pay you for defective product, nor will they pay more for good product because your efficiencies are poor. The market determines what price you can charge for your product or service; it’s your organizational makeup and internal processes that determine cost. For instance, if processes yield suspect product, the typical response is to increase the inspection frequency/sampling plan to ensure that our customers are protected from receiving bad product. That’s spending money on cost-avoidance, which is a poor ROI. That increased inspection ties up resources with non-value-added work (preventing internal failures from becoming external failures), when they’d be better utilized elsewhere, like helping prevent non-conformances from being produced in the first place. That ROI, on the other hand, is much more favorable, since it has a positive effect on things like available machine time, available floor space, reduced inspection needs, etc., which are all also benefits of Lean.

Once we realized this (which took longer than it should have, admittedly), life got easier. We stopped applying band aids and shotgun approaches and started true root cause analysis and problem-solving methods. These helped us to be able to answer some of the issues that were plaguing our poor initial test yield. Some of it was part quality, but some was simply test spec misinterpretation because we didn’t have a standardized format to follow. Still others were due to some uncontrolled manual operations on the shop floor that everyone had just assumed were being done correctly because the person doing it had done it for so long. All in all, we had lots of rocks to turn over, and had we not finally embraced tenets of both quality and Lean together, as a unit, I’m not convinced that we’d ever have solved our problems. Eventually, the management and customer calls stopped, and we were able to move on to other, more proactive things. And that was good for everyone!