5S First?

Some time ago, while speaking at a conference in the land down under, I was taken to task by a participant for suggesting, “5S is usually the first improvement” in Lean implementation.   I had carelessly adopted this posture because, as a consultant I had found that workplace organization was usually the most palatable way to demonstrate improvement on the shop floor.  (I’m not sure of this, but I think the sixth S – safety — was added at U.S. manufacturers in the 1980’s because improved safety was the only thing management and labor adversaries could agree on.)

“That may or may not be so,” my friendly heckler responded, “but just because 5S is easy, should that make it first?”

“What do you suggest as a first step to improvement?” I asked.

“Kanban,” he replied.  “A pull system is the thread that holds everything else together.”

“Pull systems are a tough place to begin,” I offered.  “Maybe it would better for a company to “get its feet wet on something less conceptually challenging.”

“No,” he shot back, “the pull system is where my company started and it’s worked very well, end-to-end.”

I hesitated, then gave a consultant’s non-committal b.s. response: “Where you start may be less critical than just getting started.  No two companies are alike,” a non-answer that ended the question, but was not satisfying either to the questioner or me.

For the next half hour, I pondered his challenge.  I recalled a quote from Hajime Ohba, then General Manager of Toyota Supplier Support Center (TSSC), about doing the right thing:

“True North is the vision of the ideal.  Always do what we should do, not what we can do.”

This seemed to contradict my “do the easy things first” theory.

TSSC had worked with my plant in the mid-90’s to help us better understand “true TPS.”  Mr. Ohba had said to me after his first visit to our site, (I’m paraphrasing) “You have made some nice individual improvements, but you will not receive the full benefit of TPS until you can put them all together.”  He asked us to identify a product line where an improvement was needed.   I picked a small assembly line that had delivery problems, one where we were not betting the farm, and we went to work.  There was no mention of 5S, or kanban, nor any lean ‘tool’ with which we were marginally familiar.

Instead there were many questions:

“Why is that assembly fixture over there?”

“How do you know what to build first?”

“Where do the finished products go next?”

“How long has this machine been down?”

One question frequently led to another.  Sometimes the answers were obvious, and other times we had no answer.

“Why don’t you have the answer?” was the next friendly but persistent question.

They wanted us to watch and think.  It was hard, and I guess it was what we should do.


Back in the land down under, about a half hour later, I caught up to my Aussie friend to apologize.  “I think I headed in the wrong direction today with my comment about 5S.  Consultants like me tend to break TPS into little pieces because it’s easier for us to describe.  Implementing TPS isn’t about 5S or kanban or any other tool.  Sometimes we just ask the wrong questions.“

“Well I still hold that kanban is first,” he replied.

“No worries,” I said.


16 thoughts on “5S First?

  1. Glenn Gertridge

    I’m not sure I agree with either one of you. In true TPS fashion, the results of the Value Stream Mapping process should determine what needs to be done first. “Picking” a starting point is purely an opinion. I believe the process is designed to let the information that is flushed out in Value Stream Mapping dictate what you work on first. The questions that Mr. Ohba asked you are things that SHOULD come to the surface in Value Stream Mapping, aren’t they?

  2. PE Moody

    I like observation as first, because a long time ago I had the chance to observe the carburator assembly line at Briggs & Stratton in Milwaukee, piles of parts, some in bins, some next to the line. I wanted to map the cycle in production because we were starting Repetitive Mfg group in APICS, which became JIT. It was pretty ugly, so then I moved into mapping the procurement cycle, which stretched on white paper all around the walls of the purchasing room – was two year cycle – no wonder it was hard to regular the material suppliers. Aluminum being delivered in a tanker every hour, awesome opportunity. But couldn’t do anything, touch any part of the system or the safety stocks, until had done a couple days of observation and a lot of questions.

  3. Jason Yip

    “do the easy things first” is still a useful principle but only after understanding what you want (aka True North).

    And understanding what is easy requires understanding what they are already good at, what resources they already have, not just assuming what was easy for us should also be easy for them.

  4. Jamie Flinchbaugh

    As it applies to 5S, I think people start here for the wrong reason – because it is easy. It is easy to implement (sometimes), but it is hard to sustain. That’s why the average life of a 5S effort is about 1 year. Here’s the problem with starting with 5S – when you fail to sustain it, then people learn unintentionally that lean is just a bunch of effort for very little gain. Not exactly the definition of lean we were looking for.

    I think one of the true motivations for starting lean with 5S is from a consulting point of view, 5S looks like progress. It might not be progress, but it looks like progress. It is impossible to argue that you’ve had an impact, because just look around you. However, there is a difference between what we call aisle lean, and true lean. Aisle lean means you look lean from the aisle. But that doesn’t mean you have the right thinking or behaviors.

    I hate to give non-answers as well, but I think the answer to this question is “it depends.” I think it depends on a great deal of inputs, and therefore is not yet formulaic, at least not yet to me. Maybe someday I’ll understand the problem well enough to codify it, but I don’t understand it that well yet.

    But here’s my formula for now. Wherever you start, it has to accomplish two things. First, it has to solve real problems that you have. If it doesn’t, then it’s window dressing, or wallpapering, or whatever cosmetic improvement analogy works for you. If it doesn’t have a real impact, people will lose interest in either your intentions or your efforts.

    Second, it has to force new thinking.The first steps have to create more questions than answers. It has to expose the fact that there is a lot to learn here. It has to create some pull; not pull of material, but pull for more learning. That’s why the participant liked starting with kanban; it was in part the right answer for them because it forced them to ask a lot of questions and take new actions.

    Where to start? This is not an easy question. And the answer shouldn’t be easy either. And as many times as I’ve gotten this question, I’ve never found the answer easy.

    Jamie Flinchbaugh

  5. Michel Baudin

    Where did your heckler come up with the notion that 5S is easy? It may look easy, but, if it really were, 5S efforts would be successful more often. The key reason it should almost never be done first is that it is so hard to make it stick. Companies that start with 5S usually have a big spring cleaning event followed by rapid backsliding that destroys the credibility of 5S with their work force.
    A consultant who recommends 5S first is like a parent telling a kid to clean up his room because he has problems at school. It probably needs doing, but it won’t solve the school problems.

  6. Pingback: 5S First? | Michel Baudin's Blog

  7. George A. Dearth

    I have started the Kanban system in my section of the company I am employed with and I find it is time consuming to start if you want all the information correct. I feel so strong about this Kanban system that my inventory quantities will drop, we’ll never run out or anything and the process can be done by anyone in our organization. It’s actually fun to set up as long as you blend this in with your normal days activities.

  8. E. Byrne

    As an introduction to LEAN, I like to start with the “Eight Deadly Wastes”.
    If a process has a problem with wait times, start with the deadly waste of “waiting”. If a process has problems with “movement”, start with that.
    This gets everone engaged right away, even people who don’t like to “sort, straighten, standardize,etc.”.

  9. Pingback: True North is the vision… | Lean Leadership

  10. Bruce Hamilton

    Thanks to all responders for adding value to this post through their insightful comments.

    A few posts back (“Illogical Progressions”) I suggested that while policy deployment is a logical starting point for the Lean journey, the organizational soil might not be prepared for that leap. Similarly, with value stream mapping, the tool may be a logical choice for recording and sharing what we see — depending upon if we are yet prepared to see; hence the title of Shook and Rother’s manual (“Learning to See”). TSSC worked in my facility for two years before showing us the VSM’s (referred to by them as ‘material & information flow diagrams’), stating, “We waited until you were ready, to show them to you.”

    I agree also that 5S is not easy at all, and that is often not only the first but also unfortunately the last improvement attempt. I didn’t intend the post to be about 5S (although I’ve noted that putting 5S in title draws a crowd:), but only about asking the right questions. It seems that any discussion of “tools”, be they 5S, kanban, VSM, etc. is likely to end in confusion between goals and means.

    1. Glenn Gertridge

      I guess my assumption was always that the “Learning To See” Lean awareness education is always first, before any activity is launched. My point was that jumping to any kind of “point solution” like 5S or any other Kaizen tool, is blindly throwing a dart at the problem. Moving forward with some sort of Kaizen activity without analyzing the entire process (Value Stream) often will cause an issue at some other individual process step, without realizing the effect. Without analyzing the entire process, and improvement in one step could cause problems with many other steps.

      1. PE Moody

        Just a general note that I love observation, a stool and a clipboard, although generally I carry a laptop to take notes – I type better than my fingers can write. You can observe for a reallllllly lonnnnngg time!

  11. Wesley Connell

    Great work with the post and comments! A great way to view some of the early barriers to implementing true Lean. Just to add a point, one of the main obstacles I have come across is the lack of the long term vision. While the CI and Lean people may be able to see the entire forest, employees on the ground floor can only see the trees.

    While it may be difficult to coordinate, taking employees to a facility with engrained Lean Thinking can allow the personnel to truly grasp what they can do with their workspace and the company as a whole. With this picture of a better way in mind, implementing new ideas such as 5S, Kanban, or VSM regardless of the order should be much more tangible to a workforce that is just learning how to see.

  12. Pingback: 5S First? « Neovista Newsfeed

  13. Pingback: Should lean start with 5S? Or somewhere else? - JFlinch

Leave a Reply to PE Moody Cancel reply