LEAN OR SIGMA ?
INDUSTRIAL MODELS
Throughout industrial history, a few firms have defined the industrial models of
their times. Modern manufacturing starts with Adam Smith’s “pin factory” and the
division of labour in the eighteenth century, which Frederick Taylor perfected as
“scientific management” at Bethlehem Steel in the nineteenth century. In the past
hundred years, taylorist productivity methods progressed to an entirely other level
through Henry Ford’s development of “mass production”. Since then, Taichi Ohno
reinvented the model at the Toyota Motor Company to deal with many of its wastes
and move it away from the “black Ford T” to encompass variety as well as
standardization.
Industrial models are often synthesized in a single catchy phrase such as “one best
way”, “production line” or “Just-in-time”, but these quick description typically
misrepresent the truly profound changes implied in each new model. Taylor’s
insight, for instance, was that by observing how the most effective workers went
about their job, work could be standardized and applied to all. To be able to do so, he
instituted engineers and clerks to realize the time-and-motion studies, piece-rate pay
to identify the most effective workers, special equipment provided by the company
as opposed to workers owning their own tools, and so on. The ultimate consequence
of Taylor’s method was to introduce into basic line organization much of the
functional complexity we deal with now.
Similarly, Ford didn’t just invent the production line. He pushed Taylorism to an
extreme by first standardizing parts (to avoid fitting), dividing labour into smaller
and smaller units, but also paying workers the $5 wage, integrating vertically, and so
on. In the same vein, Just-in-time at Toyota is not simply a matter of kanban cards,
but an entire organizational perspective involving bringing the supermarket in the
factory, continuous improvement through Kaizen, total quality with SPC,
autonomous teams, rethinking equipment for “autonomation” and so on. With
hindsight, there is a clear continuity from the “pin factory” division-of-labour model
to Toyota’s “lean” approach, as each new step solves some of the problematic sideeffects
of the previous incarnation and deals with the increasing complexity of global
competition.
COMPETITIVE PRESSURE
As their success becomes recognized in their industries, such models tend to spread
through sheer competitive pressure. Companies who adopt the new model early gain
sustainable advantage, those who don’t face an uphill struggle and eventually
disappear. This does not mean, however, that adoption is either rapid or easy. In
every case, the new model is fought hard by both workers and management. In
Taylor’s case, owners were adamant not to pay the suggested increase wage to
workers nor to invest in the necessary structure to make taylorism work. Ford’s mass
production model was portrayed as the end of artisanship in the automotive
industry (rightly enough) and ridiculed by Chaplin in Modern Times. The Toyota
model has long been dismissed out of hand as a cultural artefact by established
industrial firms.
Yet, as the success of those who adopt the new model become apparent, firms are
eventually compelled to comply or disband – a struggle in which many try but few
succeed. In the British car industry, at the turn of the XXth century, researchers
counted no less than 130 companies in Coventry alone. By the end of the 1930s, there
were only 20 independent car manufacturers left in the UK1. Adopting a new model
is rarely easy since it usually means a complete change of outlook although not
necessarily a radical change of organization. Furthermore, true experts are rare
although self-proclaimed gurus are many and as a result, reliable knowledge is not
always readily available. Finally, one should never underestimate straightforward
resistance to change, or, in many cases, management’s resistance to understanding.
In the end, the new model becomes firmly established through selective pressure as
the companies having bridged the gap drive all other competitors out of the
http://news.bbc.co.uk/1/hi/business/551543.stmmainstream market. Nonetheless, even in today’s fast-paced, change-driven society,
the acceptance of new industrial models can be counted in decades rather than years
– it doesn’t happen easily. Is it that only a few firms are capable of confronting such
changes? Not necessarily; change and continuous improvement are by and large
accepted as necessary to modern business survival and most companies are heavily
engaged in change programs – but these initiatives don’t always help them to acquire
new industrial models.
IMPROVEMENT PROGRAMS
In practice, knowledge of new industrial models doesn’t reach executives in one fully
packaged format, complete with detailed explanations and “fieldbook”. As a firm
starts to be recognized for its exceptional performance, observers identify single
aspects of a new model and make their business to convince others to adopt such
“best practices”. Bearing in mind that it is hard to grasp new vistas with old lenses, it
is hardly surprising that the first initiatives driven from early reports of the new
model are not particularly successful. The first inkling the West had of the lean
model, for instance, was through “quality circles”. These became the rage for a few
years, but soon led to disappointment and discredit – in the United States, 75% of the
Quality Circles initiatives started in 1982 had been abandoned by 19862. With 20/20
hindsight it appears fairly obvious that attempting to graft quality circles in
traditional western manufactures was bound to fail.
Still, knowledge arrives in small packets: kanban, then SPC, TPM, SMED, etc. Every
time, the consultants who make a living out of such things tend to explain that their
sliver is the necessary piece of knowledge which will resolve all the problems of the
enterprise. Total Productive Maintenance, for instance, became in several instances
“Total Productive Management” and the TPM approach was then applied to the
company as a whole. Not necessarily a bad thing, but not likely to deliver great
results either. After the spotlight moved away from Total Quality Management (the
extension of SPC to the entire company), it eventually reinvented itself in “Six
Sigma” and some companies currently claim to be “Six Sigma Companies”.
PASCALE, R. T., (1990) Managing On The Edge, New York: Simon & Schuster
Such programs have increasingly become tools of corporate management. Over the
past two decades, with increased competition in winner-takes-all global market the
recognition of a corporate need for constant renewal and improvement has emerged,
hence change programs. In the second half of his tenure, Jack Welch initiated no less
than four widespread initiatives: work-out, boundaryless organization, six sigma and
finally destroyyourbusiness.com. These are not small endeavours. According to
Welch, in 1996, the first year of Six Sigma for instance, GE spent about $200 million in
training 30,000 employees – and got somewhere in the neighbourhood of $150
million in savings3. Most of these gains appeared in service areas, processes
previously untouched by optimisation – results in manufacturing are far less
obvious.
As star companies, always the focus of media attention, invest in such programs, the
business world as a whole embraces new concepts much like fads and fashion
capture the public at large. From 1950 to 1988, Richard Pascale counted more than
thirty management fads – and this before the explosion of the nineties with Reengineering,
Empowerment, TQM, or 6 Sigma. Reengineering, for example, is a good
point in case. The concept of re-engineering companies on a core-process basis
emerged from three realizations. Firstly, the lean concept of lead-time as a key driver
to market dominance emerged in the early nineties as one of the key lessons from the
Japanese industrial offensive of the eighties, as did the notion of continuous
improvement as a source of sustained competitive advantage. Management
consultants Hammer and Champy saw an opportunity to leverage radical
improvement in lead-time by applying information technology to business processes
and thus overcome organizational “functional barriers”4. More than ten years down
the road it is still hard to assess the effectiveness of the various reengineering efforts
of the early nineties, but clearly the re-engineering movement took the business
world by storm, gathering speed in the nineties with a peak around 1994, then
moving on to public services and finally disappearing as a hot topic about 19975:
when all the hype moved on to Six Sigma on the one hand and the “new economy”
on the other. Evidence for this cycle can be backed up with citation analysis in
newspapers, articles or journals which shows a slow growth from1990 to 1992, a
3 WELCH, J. & J. A. BYRNE (2001) Jack, straight from the gut, New York : Warner Books
4 HAMMER, M. & J. CHAMPY (1993) Reengineering the Corporation, London : Nicholas Brealey
5 HOLTHAM, C. (1994) « Business Process Reengineering : contrasting what it is with what it is not », in
C. COULSON-THOMAS (ed.) Business Process Re-engineering : myth and reality, London : Kogan Page
sharp upturn in 1993 peaking in 1994/1995 and a steep plunge in 1996 (with a
significant lag in public sector articles)6. Reengineering is a clear case of taking one
idea – process focus – and applying it across the board to organizations as a matter of
ideology as much as anything else.
This is not to say that such managerial fashions don’t have an overall impact on
industry. Indeed, as sociologist Robert K. Merton pointed out, the ways in which
people define situations have real effects7. The re-engineering movement certainly
transformed how managers think about processes, much as TQM and Six Sigma have
placed fact-based and statistical approaches to the forefront. Yet, most of these
fashions pass to be replaced by the next best thing because, at company level, the
results of such programs are consistently disappointing. Beyond the much publicized
case of a few large corporate programs very few real cases of successes can be found,
and overall the economic effectiveness of these various programs is hard to prove.
Consequently, driven by the need to “do something” (and often be seen by Wall
Street in doing something) senior executives will abandon the current floundering
program for the next best thing.
THE PATH OF LESS RESISTANCE
Industrial models, on the other hand, remain relevant throughout the years. Adam
Smith’s analysis of the pin factory is just as insightful in the XXIst century as it was in
1776. Taylor’s “one best way” is a key to understanding “standardized work”, just as
Ford’s process outlook and economies of scale are still a key component of industrial
success. What each new model does, however, is not “fix” problems, but
revolutionize manager’s total perspective on industrial performance without
adopting the simplistic, one-dimensional outlook of most change programs.
Consequently, models have far longer lifetime and more profound impacts. The few
companies that seriously adopted lean wholesale as opposed to piecemeal are
thriving. Indeed, Toyota, lean’s inventor, is doing better than ever, benefiting from
hard times when its mass-production-with-added-bells-and-whistles competitors are
suffering badly.
6 JONES, M. & R. THWAITES (2000) « Dedicated followers of Fashion : BPR and the Public Sector », in D.
KNIGHTS & H. WILLMOTT (eds.) The Reengineering Revolution, London : Sage
7 MERTON, R. K. (1995) « The Thomas Theorem and the Matthew Effect », Social Forces, 74(2): 379-424
Ultimately, to compete, all companies will have to adopt lean principles simply
because it is currently the only know way to resolve a fundamental industrial
conundrum. As George Stalk phrased it, industrial cost goes down by about 15% to
25% per unit every time volume doubles, but goes up by 20% to 35% when variety
doubles8. At the moment, lean is the only proved way to increase variety and reduce
development-to-market time while maintaining low costs and premium quality. Why
aren’t more companies adopting lean wholesale rather than squander time and effort
on change programs? As any new model, lean has the same short-term advantages as
any other change program: it produces quick-wins, can be implemented on the shop
floor without having to change the entire structure and has the added benefit of
providing a roadmap to progress well beyond gathering the initial “low-hanging
fruits”. Why should companies waste time with Six Sigma (or whatever will come
next) when they could be investing in lean in the first place - and since they’ll have
to do so eventually?
The constant influx of change programs shows that modern firms can no longer be
blamed for “resistance to change”. Indeed, they now seem to be addicted to change
and only slowly coming to terms with the fact that few changes are actually
improvements. The problem with adopting a new industrial model as opposed to a
change initiative is far less a practical one than a “resistance to understanding”. New
models fundamentally challenge management’s perspective on not just specific
practices, but their very instincts and reasoning about how to run their business.
Changing one’s mindset remains one of the greater psychological challenges, and
indeed both companies and entire societies have repeatedly shown that they’re ready
to die rather than shift their outlook.
Contrarily to popular belief, changing practices is far easier than changing
perspectives – which is why programs are so popular. By now, most mass
production companies have accepted and adopted many lean aspects while retaining
their core beliefs in how business works. Consequently, they’re not obtaining the
hoped for breakthrough although they invest in one change initiative after another.
Models are fundamentally different from programs in terms of:
8 STALK, G. JR. (1991) ‘time- the next source of competitive advantage’, the state of strategy, Boston:
Harvard Business Review Press
• Outlook: Ohno’s initial insight of bringing the supermarket in the factory and
enhancing response time through short batches, levelling, just-in-time and
continuous improvement is not just a toolbox to improve the way one
currently runs the shop, it’s a fundamentally different take on where the
industrial economies of scale can be found and how to succeed on competitive
markets.
• System: consequently, a model is not simply an assortment of techniques – it’s
a system. Used independently of one another, the tools will produce local
benefits, much like Sigma or other projects will, but that’s also as far as it goes.
Used as a system however, lean tools can progressively transform operations
and lead the firm from the old model to the new.
• Roadmaps: program tools tend to be fairly one-dimensional. Six Sigma tools,
for instance, will help teams to resolve specific issues but will rarely lead them
to challenge the fundamental ways they operate and once one project is
concluded, it hardly tells them where to focus next. On the other hand, the
lean model has a roadmap to transformation which follows a step-by-step
approach from early quick-wins to more fundamental transformations.
Adopting a new model rather than a change program is not any more difficult on the
shop floor, or longer to show results, but it is harder in unexpected ways: it demands
from senior management both a considerable intellectual investment and constancy
of purpose. As one of the key developers of lean, Shigeo Shingo, points out:
“In order to successfully implement the Toyota production system [lean], you must have a
correct understanding of the basic ideas behind these principles and the knowledge of methods
and techniques to be able to implement them in a systematic way; otherwise, I fear you are
likely to make serious mistakes which will result in the failure of the system – even if you have
a clear understanding of individual techniques”9
One of the most robust psychological findings of the past century is “cognitive
dissonance”: we cling to our outlook and tend to dismiss or explain away dissenting
evidence. Changing one’s perspective doesn’t come naturally and requires the time
and effort to understand what the new model is really about, beyond the obvious
“best practices”. In many unfortunate cases, lean techniques have been modified to
fit the existing perspective of the factory, leading to bizarre consequences. Secondly,
constancy of purpose is essential because old habits die hard and instincts are
difficult to change. As managers implement the new model, they need to question
9 SHINGO, S. (1981) A Study of the Toyota Production System, Cambridge : Productivity press
every new situation in this light not to fall unawares in the old ways of thinking and
follow the path of least resistance back to square one.
All in all, adopting a new model requires ambition for the company as well as for
oneself. To many executives, change programs are an attractive alternative: they
produce quick, easily quantifiable results and don’t challenge the core thinking of the
firm’s line hierarchy. Change programs are also easy to explain and communicate.
They are a great vehicle to get people enthused and on board. Unfortunately, change
programs do not deliver in the long run. Astonishingly, after endless initiatives and
change drives, the big three automakers are no nearer to becoming lean and Toyota
continues to steadily gain market share.
SHORT TERM LONG-TERM BENEFITS
Model: productivity and quality shop
floor quick wins and opening the
minds of management and workers
Program: quantifiable quick wins in
individual or team projects which can
be done “on the side”
Model: change of perspective which
leads to a transformation of how
business is done and is a fundamental
source of competitive advantage. Also
increased barriers to competitors since
the new model is harder to acquire
Program: few long-term benefits which
leads to the abandonment of this
program and its replacement by the next
best thing
BARRIERS
Model: resistance to understanding –
as long as management has not
acquired the new outlook the
performance of the tools is
disappointing
Program: other than the cash outlay to
get things started, there are few
barriers to implementing a change
program beyond the cynicism of
experienced employees
Model: requires intellectual investment
and constancy of purpose from senior
management. To succeed in changing
the firm’s model, top management has
to get involved in more than giving a
distant blessing and signing the check
Program: no synergy from the various
spot actions which leads to
disappointing results, weariness and
ultimately disregard fro the program
A VISION OF EXCELLENCE
At the end of the day, although a lean plant is easily recognizable, no one has ever
seen a “Six Sigma” plant – nor an “empowered” plant, or yet again a “re-engineered”
supply chain and so on. Fundamentally, industrial models sponsor clear visions of
excellence in the real world of value-production. Such visions are the driving force
behind the implementation results of each new model. On the other side, most
programs tackle the side-effects of the existing industrial model, rarely proposing a
new, practical, vision. The difference is essential to the continuing progress of the
plant. As one implements lean, the factory gets transformed in practice and new
opportunities for further progress appear. For instance, a “5S” exercise will clean up
the shop floor and uncontrolled inventories will become obvious (as many other
things such as product defectives, poorly maintained tools etc.). A machine with a
stock of parts behind it is a sign of either machine reliability issues or poor quality of
produced parts. At the next level, for instance, setting up a shop stock and a kanban
production trigger will render visual the variations in production. For example, a line
with a full shop stock (overflowing somewhere else in the plant maybe?) and an
empty kanban queue is a sure sign of overlong production runs, and so forth. As the
new model is implemented, the reality of the vision of excellence is uncovered in the
factory – and guides managers in their improvement efforts. For all their worth,
programs can’t go any further than a sequence of solved unrelated problems.
Programs are only considered as serious alternatives to models as a result of shortterm
biases, in the same way that business constantly suffers the bandwagon effects
of fads and fashions which lead to inevitable bubbles and the following panics. At a
given point in time, both executives and business media tend to focus on well
publicised, simple business strategies which make sense at a given point in time but
unfailingly turn out to be flawed (think about the “new economy” and its disdainful
disregard for earnings until 2001). Programs are easy sells, both to management,
employees and investors. A new model such as lean, on the other hand, has been
around for decades, has many validated success stories on the long-run, not just a
few high-profile showcases, but is far less immediately attractive to the management
fashion industry. Ironically, most industrial programs originate in the lean model:
quality circles, SPC, just-in-time, total productive maintenance, continuous
improvement, autonomous teams and supply chain management. But, much like
pacific islanders which proudly wore western hats with their traditional costume,
rather than acquire the discipline of the entire suits, companies find it easier to pick
on some aspect or other, only to be disappointed with their performance when it
appears that new hat notwithstanding the king is still naked.
In the long run, the choice between implementing a model or a program, lean or Six
Sigma does not so much reflect on the effectiveness of the tools and techniques, but
on the management flair of the executive team. Senior managers are, as they should
be, ambitious – but are they ambitious for themselves of for their company? Pushing
one program after the other is much like picking stocks because others on the market
are buying them: although the strategy has been proved to be flawed time and time
again and leads to bubbles and the subsequent crashes, many fund managers still do
so. Investing in transforming one’s operations by implementing a new business
model is like purchasing stocks of companies after careful research and full
confidence that the price of the stock reflects the value-production future of the
company. Although all investors know this is how they should proceed, not many
can be bothered; the few who do so consistently prosper beyond any expectation.
