Managing For Lean by decree


									                                       Freddy BALLÉ & Michael BALLÉ

                                            Managing For Lean

      “There’s no other way of doing lean than the plant manager involving himself personally!” says
the manager of a Toyota plant. And indeed this seems to be the consensus amongst the successful lean
leaders, both within Toyota and outside it. Yet, although many corporate officers want to turn their
companies in “lean enterprises”, few are ready to take a plunge themselves. They’re happy to invest in
change programs, and turn over the lean transformation to lean or six sigma “experts” and their hosts
of consultants. And at first this works out fine, as the “kaizen events” pluck the low-hanging fruit and
generate expectations. But after a honeymoon of six months to one year, the program often fails to
deliver sustainable results, and senior management’s interest falters. New champions then appear to
suggest a newer, better program, to try another tack and have another go at improving quality and

      Many companies have experimented with various improvement initiatives in Total Quality
Management, Six Sigma, lean and so on, but long term results have been almost invariably
disappointing. On the other hand, a company such as Toyota which practices lean in the form of the
Toyota Production System (TPS) is moving from strength to strength, and so are the few companies
who have succeeded in creating a true lean culture. The conundrum is that although, taken piecemeal,
each of the lean tools are relatively easy to master through kaizen events and working with
consultants, mastering TPS as a system turns out to be a serious challenge. One possible reason, to
paraphrase Einstein, is that “we can’t solve problems by using the same kind of thinking that created
them.” A kaizen event will challenge a specific process, but not the thinking that created that process,
nor its day-to-day management. In truth, if we scratch the surface of the TPS tools, they all reflect a
specific way of thinking and managing – which is exactly what the Toyota senseis argue when they
claim that the real issue is to “develop a kaizen mentality within every employee.”

      Having studied lean transformation programs for the past thirty years, we believe that each of
the lean tools have deep-seated managerial implications. Furthermore, if local management doesn’t
acquire the managerial “method” underlying the tool, the gain of any kaizen workshop will be short-
lived. Thirdly, this management method is essentially based on challenging managers and applying
problem solving at all levels, in the form of systematic PDCA. Rigorously solving problems does not
come naturally to any one, and often our first instinct is to ignore the problem, sidestep it or shift the

         Freddy Ballé & Michael Ballé, 45 bd de Montmorency, 75016 Paris, France,
blame to someone else. Using PDCA and developing it in others means first and foremost
acknowledging that we “own problems” (in TPS terms, “the biggest problem is thinking that you are
okay”) and then having the persistence to tackle problems rigorously and get teams to do so as well.
This managerial behaviour involves developing a few strong attitudes about how to deal with work
situations, such as going and seeing for ourselves, challenging, involving employees, developing them
and supporting them in resolving problems in their normal operations, as opposed to fire-fighting. We
will try to show how the elements of TPS entail specific management challenges and, secondly, how
to develop the right attitude to sustain these “lean” behaviours.

      The TPS house appears in various forms of greater or lesser detail and complexity, but in its
most basic incarnation it is constituted essentially of a “roof”, customer satisfaction, two “pillars”,
Just-in-time and Jidoka, and a “base”, Standardized Work and Kaizen. The other distinctive feature of
the system, which became one basis of the “Toyota Way”, is its “respect-for-human” aspect, which
means “allowing workers to display in full their capabilities through active participations in running
and improving their own workshops.” In practice, each element of the TPS “house” constitutes a direct
management challenge beyond the more generic issue of working with the operators on improving
their own work processes.

      For instance, the goal of the system is explicitly “customer satisfaction”, in terms of highest
quality, lowest cost, and shortest lead-time. Obviously, Toyota did not create the TPS to invent a “lean
methodology.” TPS came painstakingly into being through the automaker’s constant determination to
build the best possible cars. This sort of obsession cannot come from staff functions, but must clearly
be espoused by the leadership. If, as is apparent in many businesses, senior management’s passion for
making money overshadows its interest in servicing customers, it is unlikely that kaizen workshops
focusing on delivering increased quality, lower cost and shorter lead-times will have any lasting
impact. To some extend the entry fee of succeeding in lean is an obsession for customer satisfaction
driven through the organization straight from the top. As the proverb goes “the fish rots by the head,”
and if the firms leadership is not obsessively focused on quality, cost and delivery, lean
implementation is unlikely to move beyond a cost-cutting “savings” exercise.

      Secondly, the best know feature of the TPS is its Just-in-time pillar. JIT is described by Toyota
as “making only what is needed, when it is needed, and in the amount needed.” This has a number of
practical implications for production line management. Firstly, when a car order is received, a
production instruction must be issued to the beginning of the assembly line as soon as possible. To
build on the same line different vehicles, the line must be stocked with small numbers of all types of
parts so that any car ordered can be constructed (except for a few sequenced parts). Furthermore, using
the principles of a “pull” system, the components on the assembly line must be replaced as they are

         Freddy Ballé & Michael Ballé, 45 bd de Montmorency, 75016 Paris, France,
consumed by retrieving parts from the preceding processes. This, in turn, entails that the component-
producing upstream processes constantly have available small numbers of all the needed components,
and only produce the number of parts that were retrieved by the downstream process.

      When described starkly, it appears that introducing a pull system completely changes the
management of the logistics function. In traditional setting the logistics manager handles the
information in and out of the MRP, produces production instructions, and runs the inventory systems
and manages teams of forklifts which move here and there, shifting the crates around as and when
needed. In the Just-in-time environment, logistics has to agree a stable plan with production according
to takt time, maintain the flow of withdrawal and production kanban cards (which requires a system
for the follow-up of the cards) and manages the routes of delivery trains within the factory which take
care of the circulation of small quantities of materials (small component containers) and information
(kanban cards). The challenge for the logistics manager is to disconnect the MRP for anything to do
with internal flows, and become obsessed with calculating stock replenishment loops, counting kanban
cards and the stability of the material handling routes, both internal and external (milkruns, cross-
docks, etc.). Running a couple of kaizen workshops in logistics won’t help. To sustain a working pull
system, the logistics manager will have to re-learn her job almost in its entirety.

      The second, lesser-known pillar, Jidoka, has even further reaching consequences on plant day-
to-day management. Jidoka is broadly about building quality during the manufacturing process as
opposed to inspecting non-quality out of the product after it has been built. Since inventories are so
tight, and the sequence of build order is critical to maintaining the pull system, Just-in-time won’t
work very well unless all parts moved around in the logistics flow are good parts. Hence, non-quality
must be spotted where and when it occurs, whether in a machine or at an operator’s workstation
during assembly, and dealt with on the spot. This, on the one hand, has deep implications on the
design of machines and equipments, as they need self-checking mechanisms and devices, but also on
the day-to-day management of a line or production cell. As part of Jidoka, Toyota’s assembly lines
have huge overhanging electrical billboards, which light up when operators pull an alarm chord to
show where the problem occurs. Equipment has similar warning system to notify the operators in the
cell that something is out of standard in the fabrication operations. Different musical tones are also
constantly heard in the plant to call support from logistics, maintenance, etc. when any operator is
confronted with a problem. The system is designed for immediate reactivity from management and
operators so as to avoid building bad parts. In normal operations the “andon” system keeps lighting up
every couple of minutes in a regular demand for attention of this or other workstation in the line or the
cell, as the continuous tension on the line stretches normal operations. This needs serious taking care

         Freddy Ballé & Michael Ballé, 45 bd de Montmorency, 75016 Paris, France,
      Consequences on the managerial approach are significant. Firstly, for any team of operators of
five to seven people, a team leader will be appointed. This team leader is not part of management, but
is the first person in the system to react to an operator facing a problem, the first line of defence.
Either the team leader can quickly identify the problem and with a word or two sort it out with the
operator, or the line will eventually stop, and a group leader come to help with finding a
countermeasure. Andon chords pulls are carefully analysed to understand which workstation creates
the most problems, under what conditions. Now, in most traditional factories, the supervisor
(equivalent of the group leader) spends most of his time fighting fires, which is trying to fix the line
when it’s been down for a good while, looking for missing parts, or shifting operators around if one
cell is blocked for machine failure, missing parts, or sudden change of production plan. Support
functions are used to react in their own time, according to their own priorities and established pan.
Imagine the impact reacting immediately to the calls of an andon system can have on a shop floor
management not explicitly structured to deal with it. And in real life, this is what happens: Just-in-time
experiments fail because the line can’t be stabilized enough to cope with Just-in-time stock conditions,
because day-to-day management can’t be geared to the kind of reactivity required by Jidoka and
building quality into the product.

      Finally, the “base” of the TPS house, standardized work and kaizen is no less demanding from a
management method perspective. The first big shift away from traditional management is that
supervisors are held accountable for safety, productivity and quality results. They’re asked to track
carefully the accidents in their area, the line stops and their causes, and the quality issues encountered
by the team members. Supervisors have to work with the teams to come up with plans to improve their
shift’s performance on a routine basis. This is a genuine departure from the typical supervisory role of
making sure enough operators have been scheduled on the processes to produce the required
production instructions and that big problems that keep cropping up are contained without too much
collateral damage. These performance improvements are expected through a rigorous application of
existing standards, which involves training operators to standards and their use, as well as generating
ideas to improve the standards. Indeed, the suggestion system is intricately linked to standards
application since suggestions are supposed to respond to problems, themselves defined as a gap
between current situation and the standard. It’s the supervisor’s job to maintain standards, train
operators to the standards and encourage them to suggest practical ways of keeping to the standard
when its is hard to reach continuously for whatever reason. The suggestion system is also a powerful
device for making sure the supervisors actually listen to their operators. Again, the management
method involved with maintaining standards and sustaining a suggestion system is pretty explicit, and
often radically different from the systems in place in non-lean companies. Standards, in particular, are
at the heart of “visual management”, since any aspect of the factory should be visualised so that
anyone can see at a glance whether the situation is within normal conditions or not. Visual

         Freddy Ballé & Michael Ballé, 45 bd de Montmorency, 75016 Paris, France,
management has little to do with powerpoint indicators posted on whiteboards. It is mostly about
simple practical systems such as production Analysis Boards, Red Bins, markings which tell you very
simply where something is going on that requires immediate management attention.

      Last but not least, the “respect-for-human” dimension of the system also implies that
supervisors, supported by their hierarchy, do what they can to help with operator’s personal and
professional problems. Just this, in itself, is a radical departure from the way most shop floors are
routinely managed. First, this means that supervisors are trained to handle such discussions with
operators. Second that they have clear guidelines about which problems they can help with, which
they won’t try to, and which need to be referred to the hierarchy. Thirdly, supervisors need to be
equally supported by the rest of the management structure, all the way to the plant manager, to be able
to do this safely. Clearly this approach to human relations has yet more staggering implications to the
management’s attitude towards, well, management.

      At the end of the day, to sustain lean outcomes, one needs to imbue one’s management structure
with a passion for servicing customers, to create the kind or requisite shop floor organisation to react
to an operator call every other minute, to ask from line management continuous safety, quality and
productivity results through the development of work standards, and to build a suggestion system
which will involves operator’s minds as well as the job uses their arms and legs. To say the least, none
of these management behaviours are what most managers have been trained for. Re-training line and
functional management to lean practices is an arduous effort which is typically achieved in lean
through sensei visits on the shop floor, and the participation of managers to shop floor activities such
as kaizen, SMED or 5S workshops. Equally importantly, senior management must keep this training
drive focused over time and demand from its management a continuous stream of activities to achieve
budgets. Typically, as managers undertake a number of activities themselves, and under the guidance
of a sensei who ensures the problems are well identified and the principles well understood, they will
start to see the links in the system and then leverage their understanding to truly apply continuous
improvement. There is no such thing a plant finally reaching a state of “lean”. Lean is not a status, but
an outlook in which waste is eliminated and problem resolved, continuously.

      Why would functional managers be ready to question and change so fundamentally the way
they do their job? And have done it for as long as they’ve held the post. Why should they think they
need to change? Why should they accept the lean solution as the right thing to do? And why should
they risk taking such radical steps such as unplugging the MRP, having to respond to every operator
doubt, drawing out standards instead of letting manufacturing engineering do it and so on? How can
they be helped to face the risk of such radical moves without the support of hundreds of TPS trained
coaches to help iron out issues? The Toyota senseis have an answer; they consider that TPS is nothing

         Freddy Ballé & Michael Ballé, 45 bd de Montmorency, 75016 Paris, France,
else than the application of scientific thinking to production, and that lean techniques will arise as the
result of applying lean thinking one step at a time. In this sense, the scientific method applies to the
shop floor as rigorous testing of hypotheses on well-defined problems. Ultimately, they consider that
if one tries hard enough to reach customer satisfaction through zero defects produced, building parts
one at a time and in sequence on demand, and with the most value added possible, taking away all the
waste, one will end up with “lean” solutions – which are likely to be similar to every other lean
solutions, but relevant to the local application. On the other hand, forcibly applying lean tools to badly
understood situations is likely to increase confusion more than anything else.

      Which is the heart of the matter: one has to really, really want to do this. Firstly, the
management team has, one by one, to acknowledge the competitive gap and want to do something
about it. Secondly, they have to learn to apply the lean principles to their own specific situations.
Thirdly, they have to be ready to deal with all the confounded nitty-gritty problems applying lean
principles will generate or reveal. Fundamentally, the essence of TPS thinking is to be willing to start
small, learn through quick trial and error by responding to specific problems, and keeping at it, rain or
shine. So rather than applying cookie-cutter solutions to badly understood problems, the real issue is to
encourage in everyone an excitement about problems, and a persistence about resolving them —
problems can be seen as opportunities to improve. This is not a trite formula. It’s a prerequisite of
embarking on the lean journey. Over the years, we have observed many non-Toyota factories trying to
go lean, with varied degrees of success. Overall, the ones, which succeed are the ones where the
management catches the TPS-like obsession with solving problems in normal operations, and resisting
the siren calls of being heroes by glorious fire fighting. This is not to say that fires can be ignored!
Customer satisfaction comes first, and of course plant managers are expected to get all urgent issues
solved, quick – but not exclusively. The further observation is that this problem-solving bend can be
fostered in a management team and all the way down to the shop floor by practising simple, but
demanding attitudes towards work situations. These attitudes involve: going and seeing, challenging,
involving, developing and supporting.

      “Go and see for yourself,” is one of the key principles of the Toyota Way, Toyota’s statement of
its fundamental values. The idea is never to resolve, or even discuss, an issue without being at the real
place (where it occurs), with the real people (who do the work on a day-to-day basis) and the real
facts. Although this might seem mundane and simple common sense, it is a profound transformation in
behaviour. Since Alfred Sloan demonstrated in General Motors that a huge corporation could be run
effectively “by the numbers”, simply by making decisions based on reporting of financial indicators,
the “command centre” myth of management has taken a firm hold on the corporate mind, supported by
the cornucopia of “decision” software. Armed with the best executive systems, managers dream of
seating at their desks with all the levers in place, and single-handedly run their corporate empires.

         Freddy Ballé & Michael Ballé, 45 bd de Montmorency, 75016 Paris, France,
Lean’s approach is radically different. Decisions should be taken at the point of value creation, where
work happens. This does not preclude from using analytical software to inform decisions, of course,
but this sends the decision-maker always back to the place where “things are made.” For instance, the
project manager for a new American car will spend the time to drive with a core team through all the
states of the union to get a gut feel of US driving conditions. The team will then visit car dealers and
plants to see for itself how the new car could be sold and built. Then it will propose its strategy. Mr.
Fujio Cho, the ex-president of Toyota, could often be found on the shop floor observing operations
and wondering how to improve conditions. “There is a secret to the shopfloor,” said Taiichi Ohno,
one of the inventors of TPS, “ just as there is a secret to a magic trick. Let me tell you what it is. To
get rid of muda [waste] you have to cultivate the ability to see muda. And you have to think about how
to get rid of the muda you’ve seen. You must repeat this- always, everywhere, tirelessly and

      Most managers are focused on solving large, difficult problems, and using extraordinary means
to solve extraordinary situations. The assumption underlying “go and see” is that we should focus on
normal conditions, see the issues which appear as waste, overburden of operators or machines, and
stop-and-go policies. A gain of 4 seconds on an 18 seconds cycle by discussing with an operator at a
press about how to use his two hands instead of one is a twenty percent gain! Imagine if this can be
achieved on every workstation – and it usually can. Most managers would have simply walked past
this cell because there is no fire to fight. However, managing with lean will make you stop and look,
discuss with the operator, and find a way to improve his working conditions and the quality and
productivity of his work. Go and see.

      In lean, the biggest problem is thinking you are okay. Continuous improvement is spurred by an
attitude of constant challenge, which expresses itself in three main ways. First, giving oneself stretch
goals: objectives difficult to reach, but not impossible. The way to do so is to plan as if all went well,
taking the slack out. In this way, the goal is to strive to make each part of the system or the project
work at its maximum capacity, without being discouraged by an impossible “artificial” goal. On the
shop floor, for instance, the target cycle time is taken by the minimum of five cycles of the best
operator. If people follow the standard method and nothing comes to interrupt them, the goal is
achievable – it’s stretched, but not impossible. The second aspect of challenging oneself is not
accepting to dismiss failure to resolve a problem because of external factors. Sure, customers are
unrealistic, overdemanding and change their minds all the time. Certainly suppliers are unreliable,
inefficient and can’t provide the required quality, and so on. But still, within each problem, the
question is: “how could I have handled this better?” If capacity is an intractable issue, what could we
have done about containment? And so on. Challenging oneself is not natural behaviour and needs to
be practices every day – it’s a mental habit, which needs to be deliberately acquired. Lastly, teams and

         Freddy Ballé & Michael Ballé, 45 bd de Montmorency, 75016 Paris, France,
collaborators also have to be relentlessly challenged. Because they’re uneasy with conflict, many
managers walk past an abnormal situation and don’t say anything - until evaluation day comes, or
there is a big problem, and then the long list of grievances is thrown all at once at the employees’ face
(who, not unreasonably, is likely to dismiss the whole thing as unprofessional or unreasonable).
Challenging means confronting every abnormal situation on the spot. The shop stock has an empty
lane? Discuss with the supervisor why this is so. A bad part is routinely placed into the red bin without
notifying or analysis, discuss with the operator why this is so. The workstation’s housekeeping is
disappointing; discuss it with the team leader. Much of the challenging attitude is not to walk past
abnormal situations without systematically pointing them out. “Fix the problem, don’t fix the blame,”
say the sensei. This is not about giving people a hard time, but teaching them to see problems and to
resolve them routinely. As Eiji Toyoda once said: “problems are rolling around in front of your eyes.
Whether you pick them up and treat them as problems is a matter of habit. If you have the habit, then
you can do whatever you have a mind to. You don’t have to look for them. You just pick them up.”

      The third aspect of managing for lean is people involvement. It is hard to expect staff to react
positively to being challenged if they’re not involved, if they don’t feel there is an issue is the first
place. For instance, developing lean approaches in hospitals is generally fairly easy because
employees are trying desperately hard to care for patients, and are frustrated by all the problems they
encounter. Getting them to get to grips with problems is a matter of breaking them down in
manageable parts and coaching them to resolve issues piecemeal. In other environments though, no
one feels there is any particular problem. “No problem is a problem,” as the saying goes, because it’s
going to be hard to challenge and improve if no one is involved! Involving people in managing for
lean has a very specific sense: it means sharing our problems with operators. The usual way is to get
them to track their quality, productivity and delivery performance, through “hourly production
analysis boards” to start with. By tracking their own performance, employees will have a basis for
discussing the need to improve and identifying the issues that cost them most in terms of performance.
There might then be many discussions on how to improve, but not on the need to improve.
Involvement starts with explicating performance, so that at the very least people are clear on what is
expected of them, and then sharing more general issues and concerns, so that all understand what the
management is trying to do and the direction the company takes.

      Involving is the first step to developing people, the core of managing for lean. Again, the
meaning of “developing” is very specific in the lean context: getting people to rigorously solve
problems. The assumption is that employees have been developed if they learn technical things about
their job, which allows them to do it better. How can they learn best? By resolving specific problems.
Problem resolution, the PDCA, is taken here in the sense of the scientific method – quick

         Freddy Ballé & Michael Ballé, 45 bd de Montmorency, 75016 Paris, France,
experimentation. In this sense, the PDCA cycle is a management method in itself, to steer people
through the systematic resolution of issues.
                 • To start with, why is this a problem? And why is this a problem worthy of our
        attention? A good understanding of the background of the problem helps with stating exactly
        what problem we are trying to resolve. In lean, a problem is defined by a gap between current
        situation and standard, so problems can be described in very narrow, specific ways.
                 • What results do we expect? Stating the goal of a problem is very important because if
        we don’t, we’ll never know whether we have actually resolved the issue, and learned
        something. For instance, it is easy to be satisfied with a vague “improvement”, simply by
        looking at the situation and gathering the low-hanging fruit. A challenging goal is necessary to
        challenge the actual system, which produced the problem.
                 • What is causing the current situation? Analysis is about explicating clearly first, the
        point of cause – where does the problem actually occur, in the spirit of “go and see.” Then it is
        important to try to list the contributing factors, and, like with a scientific investigation, to run
        quick experiments to see whether the identified factors actually affect the point of cause. For
        instance, before going into complex DOE reasoning on problems with an injection press, one
        could decide to pay particularly attention to clean the problematic mold for a week to see
        whether cleanliness is a relevant factor. Such quick experiments allow the problem solver to
        learn about the real place and real facts, and to identify concretely which aspects of the
        situation really count. Finally, once the correct factor has been identified, a “5 why?” analysis
        is the way to understand what contributed to this factor being affected, and leads to the root
                 • What are the specific actions to solve the problem? And specifically how do lean
        overall principles (customer satisfaction, 0 defect, 100% value-added, producing parts one by
        one in sequence with customer demand) apply and what kind of specific solutions can be
        found? Also, what is the intended effect of these solutions? In other words, who will do what
        by when, and how do we intend to manage this implementation plan.
                 • How do we know if these actions are having an impact? One of the greatest drawbacks
        of the “workshop” strategy is no one is left to care for the problem after implementation. In
        lean the problem owner is supposed to spend all his or her time after implementation, to make
        sure the impact is what was intended. If the results are lower than the goal, fundamental
        questions must be asked: the odds are that deep-seated assumptions have not been challenged
        hard enough.
                 • What else should be done? How to we obtain the remnants to get to the original goal?
        What has been learned in technical details? Would any one else benefit from this learning, and
        how can we involve them?

         Freddy Ballé & Michael Ballé, 45 bd de Montmorency, 75016 Paris, France,
      Driving the PDCA relentlessly is the management method. As they do their routine jobs, staff
will be asked to continuously handle quick PDCAs, and managed accordingly. It is easy to envisage a
situation where everyone in the organization is working, at any given time, on a specific PDCA. If this
is achieved, the lean approach finally stands on its own two legs: getting results by continuous
improvement and developing people by resolving problems.

      Finally, although there is a strong bias in lean for getting people to acknowledge they have to
deal with their side of the problem first, if only to establish credibility with other partners on the basis
of “sweep your on front porch first,” sooner or later teams will run into issues they cannot resolve on
their own. Supporting them on those points is essential to establishing trust relationships and
motivation. For instance, clerical employees can do only so much to reduce the variation in their work
without being able to touch the computer programs they use throughout the day. Resolving IT issues is
way beyond their ken, and requires a management action of IT resources to fix the problems there.
The implicit social deal in lean is “solve your problems and I’ll help you to improve your working
conditions.” In too many cases, employees hold their own of the bargain, but management reneges,
pleading that the problems are simply too complex. This argument in not receivable in managing for
lean. If the software causes difficulties it is specifically the job of management to deal with the
software provider or the IT department to solve the issue. “Challenge” is a deep-seated attitude which
applies to all, regardless of organization position or posture.

      Going lean cannot be delegated or purchased. Lean transformation is about knowing what you
want, and developing a genuine, honest obsession with customer satisfaction. Indeed, by construction,
lean is a challenge for the line hierarchy, from the CEO to front line management, not for staff
“experts.” Doing so requires a fresh look at basic management attitudes towards management itself.
By observing many attempts at lean transformation we have tried to formulate the basic management
practices to sustain a successful lean approach. Beyond keeping a deep concern for customer
satisfaction as the “true north”, managing for lean rests on walking on two feet: achieving results by
solving problems in “normal conditions” – every customer matters, every part counts, every cycle
counts – and developing people through involving them in rigorous PDCA. To do so, the line must
develop, through constant practice, five underlying attitudes: “go and see,” challenge, involve, develop
and support.

         Freddy Ballé & Michael Ballé, 45 bd de Montmorency, 75016 Paris, France,

To top