‘Tis The Season for Management by Measurement

It must be that time of the year. I see traffic in the online forums asking about how to set key performance indicators for lean staff people so their performance incentives can be set.

If anyone were to ask my advice here, it comes down to one word:

Don’t.

Two reasons.

First, we have overwhelming evidence that these incentives not only don’t work, but they actually make performance worse in the kinds of work you are asking these people to do.

The Overjustification Effect

Dan Pink on motivation

All of this is consistent with what Deming told us decades ago, yet we keep doing it.

The second reason is inconsistency and misalignment.

Having the continuous improvement staff operating to separate metrics disconnects their efforts from line management’s. They become responsible for improving the operation while the line management processes are… what?

The message to the shop floor team is pretty clear here. They can read an organization chart. Do what the boss says, then, if we have time, and if the improvement guys can make the case, then maybe listen to what they have to say.

If you must have management-by-KPI, then the performance measurement for continuous improvement must be exactly the same as the line manager being supported. Why?

The question I would ask is “Who is responsible for the performance of the organization?” If it isn’t the line leader, then why does that position exist?

It makes no sense whatsoever to have the lean implementer working to a different agenda.

Our management traditions of de-aggregating and delegating are not serving us well. We need to take a systems view and realize that everything is inter-related. Further, we need to grasp that B.F. Skinner’s (dubious) research on rewards-based-behavior simply does not apply to management. Never has. Wishing otherwise isn’t going to change it.

Decisions, Decisions

How many “If-Then” steps do your team members have to deal with in the course of their routine work?

Every one of those branch points is a decision. It is a point where the team member must memorize decision criteria and the correct choice(s).

Each “If-Then” in the process flow potentially doubles the number of possible paths the process can take.

Each decision is an opportunity to make a mistake.

The more complex a process, the more time and experience the team member requires to master it.

Mental bandwidth is limited.

The more attention they must expend to do it right, the less they can devote to thinking about how it could be done better.

How complicated a world do you create for people trying to do the work?

The more “flexible” your human interface with the process, the more complicated it is for the person who has to use it.

Do they have to enter ad-hoc query criteria into computers to pull information they routinely need every day?

How many decision criteria are things that people “just know?”

How often does someone encounter a problem or new situation and get a verbal instruction from the supervisor on how to handle it? What happens then? Maybe a general announcement at the next team meeting, if you’re lucky?

Go down to your work area.

Watch how people interact with the routine work.

Each of those decision points is an opportunity to simplify your process flow and make life a little less stressful for all of you.

The Tough Decision: What Not To Do

Today’s Dilbert strip highlights a situation that is only funny because it happens so often:

The idea that a company can focus on 25 key areas, or 125 key performance indicators (yes, I said 125 because I have seen it myself) is obviously ludicrous.

Of course a manager has a legitimate concern to ensure people don’t take their eyes off things that are important to focus on something else.

But the leader’s role here is to ensure there are processes and systems in place that anchor the routine things. Further, those processes and systems need to be designed to alert the appropriate people when something goes out of control, or past a boundary.

Without having any routines the manager has no choice but to make everything something to focus on.

Because there is no standard process, everything must be managed as an exception.

This stresses the organization because in reality they can only micro-manage a few things at a time. It is left up to the people to decide what isn’t going to get done. The bosses response at that point is going to be negative no matter what they accomplish. “Respect for people” does not play here.

Leader’s toughest job is deciding what we are not going to work on right now. It isn’t as tough as it seems because if a focus or challenge is selected well, it actually organizes the problem solving effort to pull just about everything else in behind it.

Let’s take an example from a previous post: On time delivery.

If we say “We are going to emphasize “on time delivery” as a theme or challenge this year, what kind of things might people end up working on?

  • Having every operation start on time.
  • Sources of delay.
  • Quality issues (which cause delays)
  • Safety issues (also cause delays)
  • Visual controls and good response to problems (to get on top of problems quickly)
  • Equipment reliability.
  • Setting a good operational takt time.

The list goes on. But by having a decent challenge, the local area can focus on the things that are impacting their ability to deliver on time. It isn’t necessary to make a laundry list of everything that could cause a delay and measure it from the top of the organization. If you do, everyone’s energy is dissipated working on problems that they don’t necessarily have.

Of course this is all based on having a leadership process that gets down to the work area, grasps what people are actually working on, and coaches them through the process of aligning their efforts and solving the right problems in the right way.

Hmmmm… so does that mean that the #1 thing to focus on if you want consistent on-time delivery might be leadership development?

I guess I need to get back to the Liker and Convis book.

Once Again: What Doesn’t Work

The introduction of The Toyota Way to Lean Leadership covers ground that:

  1. Has been covered before – we know all of this.
  2. Needs to be covered again, because most people act as though we don’t know it.

Simply put, Liker and Convis (legitimately) feel the need, once again, to let us know the things which reliably fail when trying to build a sustaining culture of continuous improvement.

“So let’s train some Lean Six Sigma experts to grab the tools and start hacking away at the variability and waste that stretch out lead time; this will make us more successful, both for our customers and for our business. What could be simpler?

What indeed.

I am, once again, reminded of a saying that “People will exhaust every easy thing that doesn’t work before they try something difficult that will.”

The authors cite some of the same things that we have heard before:

  • Trying to determine ROI for each individual process step, or each individual improvement, doesn’t work.
  • Trying to motivate the right behavior with metrics and rewards doesn’t work.
  • Trying to copy the mechanics doesn’t work.
  • Trying to benchmark, and copy, a “lean company” in your business doesn’t work.

Yet, even though we have been hearing these messages for at least a decade, actually longer, I continue to encounter managers who try to work this way.

The authors assert, and I agree, that this is the result of people trying to fit Toyota’s system into a traditionally taught management paradigm that is so strong people aren’t even aware that there is a paradigm, or can’t conceive there is anything else.

They are stuck inside their threshold of knowledge when the answer is beyond it.

This reminds me of the Edwin Abbot’s 1884 story of Flatland, a two-dimensional world populated by creatures who cannot conceive of “above” and “below” their planer existence.

Although his story is often read by students struggling to grasp models with four, five and more dimensions to them, it is really a story of social change and paradigms.

Our management systems are a “flatland” with Toyota’s system existing in a space that we have to work hard to grasp. We can see pieces of it where it touches ours, but like the creatures in Flatland who only see two dimensions of three dimensional objects, we only see the pieces of TPS that we can recognize.

Support vs. Involvement

I spent last week teaching three sessions of Job Instruction. One session at the end of night shift starting at 5am, the second session catching the start of day shift at 7am, then the third session for swing shift at 1:30.

What is cool, though, is there is a senior member of the site leadership team in every session. They are not just observing, they are active participants.

Next week those leaders, with input from the team members in the class, are going to caucus and discuss the ramifications and how to deploy this method across their operations.

This is critically important because just teaching a few dozen people a skill and expecting them to each apply it independently doesn’t work. The company has to also develop the ecosystem that it can flourish in.

But this has been the pattern from this company from the beginning.

The Operations Manager has been leading the operations kaizen teams, working hard along side the team members to develop disciplined, effective pull systems. This is a high-variety product mix, with lots of different routings, so there were some real challenges to overcome. I was really impressed with their application of what would normally be very advanced thinking.

Results? Over the last six months, their metrics have taken a clear turn upward. You can see the inflection point on the graphs. Their company sister operation is starting to take notice. Good things all around.

A plant manager once shared a conversation he had with his CEO. The teaching is important, but what makes it work is willingness to be taught.

When we talk about management involvement, this is the crux. This team has been curious, willing to try and explore things – in front of their team members – and willing to keep at it until they figure out how to make it work.

That’s what it takes.

Mike Rother: Time to Retire the Wedge

Note – this post was written pretty much simultaneously with a post on the lean.org forum.

Mike Rother has put up a compelling presentation that highlights a long-standing misunderstanding about the purpose of “standards.”

Some time ago, a (well-meaning) author or consultant constructed a graphic that shows the PDCA wheel rolling up the incline of improvement. There is a wedge labeled “Standards” shoved as a chock block under the wheel to keep it from rolling back. That graphic has been copied many times over the years, and shows up in nearly every presentation about PDCA or standard work.

The implication – at least as interpreted by most – is that a process is changed for the better, a new standard is created, and people are expected to follow the standard to “hold the gains” while they work on rolling the PDCA wheel up to the next level on the ramp.

Slide 6 (taken from the book Toyota Kata) shows the underlying assumptions that are implied by this approach, especially when it doesn’t work.

There are some interesting assumptions embedded in the “wedge thinking.”

The first one is that “the standard can be ‘held’ by the people doing the work.

That, in turn, means that if when things start to deteriorate, the workers and first line leaders are somehow responsible for the “slippage” or “not supporting the changes.”

With this attitude, we hear things like “Our workers aren’t disciplined enough” or “How do I make them follow the standard?” The logical countermeasures are those associated with compliance – audits focused on compliance, and sometimes even escalating punitive actions.

Back in my early days, I had a shop floor team member call us on it quite well: “How can you expect us to hold some kind of standard work if the parts don’t fit?” (or aren’t here, or the tools don’t work, or jigs are misaligned, or the machine isn’t running right, or someone is absent, or we are being told to hurry and just get stuff out the door?)

This is the approach of control. The standard is fixed until we decide to change it.

Taiichi Ohno didn’t teach it this way.

Neither did Deming or Juran. Neither did Goldratt. Nor does Six Sigma, TQM, TPM.

Indeed, if we want creativity to be focused on improvements, we have to look up the incline, not back.

We are putting “standards” on the wrong side of the wheel. Rother’s presentation gets it right – the “standards” are the target – what we are striving to achieve.

The purpose of standards is to compare what we actually do against what we wanted to do so we know when they are different and so we have some idea what stopped us from getting there.

Then we have to swarm the problem, and remove the barrier. Try it again, and learn what stops us the next time.

The old model shows “standards” as a countermeasure to prevent backsliding, when in reality, standards are a test to see if our true countermeasures are working.

I believe this model of “standards” as something for compliance is a cancer that is holding us back in our quest to establish a new level of understanding around what “continuous improvement” really means.

It is time to actively refute the model.

If you own your corporate training materials, find that slide (it is in there somewhere) and change it.

If you see this model in a presentation, challenge it. Ask what should happen if something gets in the way of meeting this “standard.”

“What, exactly do you expect the team member to do?”  That sparks an interesting conversation which reveals quite a bit.

“Find the Bright Spots”

One of the problems facing all of us – from pundits to practitioners alike – is “too much information.” We look at a complex state, like the way Toyota operates, try do describe it in great detail, break it down, build models, and say “OK, make it look like that.”

So one of the most common questions is “Where do I start?”

The platitude is “start with 5S” but in reality, that doesn’t really work very well either. It doesn’t really drive a cultural and behavioral shift. If you have to audit people into compliance, that’s how it is working for you.

What we know today is that TPS is much less about how it looks than it is about what people do.

I have seen a handful of other companies who have managed to get a true continuous improvement culture running (at least for a while). There is something very different about them vs. a standard “lean implementation.”

Yet these companies have the same caliber of people, the same resources, the same baseline problems as everyone else. They operate in the same environment, and yet operate differently.

A key point in Switch is “find the bright spots.” That is – look at who has success in the same environment, and grasp what few factors are actually making a difference.

Perhaps, rather than “looking for waste” we ought to be looking at what few things make a big difference. Just a thought.

Why Don’t They See This Is Better?

“Resistance to change” is a common theme of discussion among practitioners on various online forums, as well as in emails I get from readers.

One thing I see fairly often is that a practitioner will be suggesting a visual control or a specific application of a “lean tool” as a “better way” in the process being examined.

“They can just look it up on the computer,” say those holding on to the status quo, “why do we need to put up a board?”

Why indeed?

So the practitioner tries to make a logical case, and often comes away frustrated. “Leaders aren’t supporting the changes” is a common lament at this point.

But let’s break down the problem and see if there is more we can do.

We are often debating whether or not a particular solution is better than the current way.

But in our “implement the tools” approach, we tend to make “lack of a specific solution” into a problem.

Whoa. Let’s back up a bit and see if we can head this off.

Do you have agreement on a clear target objective, one that all parties can describe? Do you know how the process should be performing?

Note I said “should” not “could.”

“Could” is potential.

“Should” is an unmet expectation. Big psychological difference there.

If everyone agrees that the status quo isn’t getting it done, and also agrees on what they want to achieve instead, then the next question is “OK, what is stopping us from taking the next step?”

This shouldn’t be an abstract exercise. As you watch the people in the process try to reach a higher performance level, look for “What just got in our way?”

You need to help the leaders, and your other constituents see it with their own eyes. Don’t expect them to take your word for it. You wouldn’t take theirs without your own observation.

If everyone can see, for example, that a team member gets too far behind to recover before anyone else notices, or that a machine is experiencing stoppages or excessive changeovers, for example, then you can start discussing solutions.

Perhaps the team leader needs to make quick status checks periodically, in a way that is not intrusive.

What is stopping him?

Well, that’s difficult right now, because everything is buried in the computer, and often updated in batches after the work is done.

Hmmm.. What could we do to make things more visible, in real time? Is there a way we can set up the work area so the team leader (and the worker, and anyone else just happening by) could readily see there is an issue here?

Now, and not before, is the time to start discussing solutions. But you can’t just make the logical argument. You have to get agreement each step of the way.

That might very well take longer than you want it to. People are funny that way.

But the bottom line is this: “Lack of your pet solution,” no matter how many books and name-brand authors refer to it, “is not a problem.”

We create a lot of our own resistance by running into things, and leaving fires behind us.

5S in Three Bullets

I was in a conversation today and we ended up boiling 5S down to three key points:

  • You have everything you need.
  • You need everything you have.
  • You can see everything clearly belongs where it is.

Of course at the next level, these statements are the standards you are continuously checking against.

Presumably we have cleared out everything else, leaving only what we thought was needed, and established visual controls to verify we have those things, and only those things, in the work area.

Then, as the work is done, the moment someone discovers something else is needed, THAT is the time to deal with the issue.

– Ask “Is this something we should need in the normal course of the work?”

If so, then you learned something that you didn’t know or didn’t remember when you first organized the area. Add that item, find a place for it, and establish a visual control. Right now.

If not, then “Why did we need it this time?”

What broke the normal pattern of work?

This is where 5S breaks down – when we don’t discriminate between something that is needed in the normal course of work, and something that is needed as an exception.

If we just “get it” and add it to the work area, then we normalize deviance and incrementally erode the process. If we ignore the issue, we add “getting this when it is needed” to the work cycle.

If, on the other hand, we seek to understand what broke the normal pattern and deal with the core issue, we have a shot at real kaizen. (It is perfectly OK to get what you need and keep it around as a temporary countermeasure. Just put it someplace where you will KNOW when you used it.)

The worst thing you can do is allow these small problems to accumulate and try to correct them en-mass as some kind of “corrective action.”

Kanban

Likewise, kanban can be expressed the same way. It is more dynamic, but is really answering the same questions in the context of materials.

 

Standard Work

If you paraphrase these key points to just about any other “tool of lean” then the purpose of surfacing problems and driving solution becomes apparent.

  • You are doing everything that is required.
  • Everything being done is required.
  • Everything being done clearly is part of the sequence.

Take a look at the other classic “tools of lean.” How would they fit into the same pattern?

 

Toyota Kata Handbook

Mike Rother has made some significant revisions to his Improvement Kata Handbook.

 

  • The role of “True North” is much better defined as the context of improvement.
  • He has filled in a lot of valuable detail for “Grasping the Current Condition” and setting Target Conditions.
  • The structure for the PDCA cycle has been tightened up.
  • And the time, place and method of coaching is much more explicit.

Even if you took a look earlier, get the latest and study it.