Jim Womack shares a future-focused, humble reflection of the groundbreaking book he co-authored with Daniel T. Jones and Daniel Roos, The Machine that Changed the World, during its 30 Anniversary.
Roberto Priolo: The Machine that Changed the World is considered a classic. But how was it received when it first came out?
Jim Womack: At that time -- it was the fall of 1990 -- there was no rapid feedback from the publishing channel. Because publishers’ ability to predict demand was so poor, they sent an enormous number of copies of books they expected to sell to distributors and, until that first printing was exhausted, we had no idea how well a book was doing. So, there was a long silence after Machine came out, compounded by the fact that the first Iraq War was happening at the same time.
Something that probably helped us was that the U.S. economy was going into recession, and General Motors was tottering on the brink of bankruptcy. Machine argued that mass production companies were under attack by lean producers, and by the spring/summer of 1991, it was pretty apparent that we had said something that resonated with people in the U.S. and Europe. In the end, we never knew how many copies were sold, but we noticed that our publisher started to sell the rights for the book for translation in many languages -- which was clearly a good sign. My personal metric for our success was noting how many people recognized me at the Detroit airport, which went from zero in September 1990 to a lot by the spring of 1991.
The academic response was lukewarm at first. Our investigation wasn’t typical academic practice: rather than send out surveys and questionnaires, we went to the gemba to observe the actual work. And rather than present a grand theory from the outset, we worked backward from what we observed. We didn’t publish through the academic press, either, so perhaps the academic world saw Machine as a conscious attempt to “jump the fence.”
RP: As you set off researching the system in place at Japanese automotive companies compared to Western ones, did you already know you’d be stumbling upon something so revolutionary for the business world?
The Virtual Lean Learning Experience 2020 (VLX), hosted by the nonprofit Lean Enterprise Institute (LEI), features a two-week immersion into LPPD principles and practices. Register and hear from the foremost practitioners and coaches how you can design your company's future with LPPD. For more information about:
- participating in a FREE VLX 2020 session featuring Jim Womack and Jim Morgan
JW: Between 1979 and 1984, we had worked on a project that led to the book The Future of the Automobile, which in broad-brush strokes described much of what is in Machine. The problem was that we didn't have any credible performance data nor a way to link performance with methods. So, we knew the Japanese (in particular, Toyota) were much better from the very beginning; we just couldn’t explain why in a way that would be widely accepted.
The prevailing theory in America and Europe as we started was that there wasn’t a level playing field. The Japanese were cheating by paying their second- and third-tier suppliers peanuts to do the bulk of the work, they were artificially weakening their currency, and government and industry (“Japan Inc.”) were working together to target foreign market segments for capture. This idea was so widespread that when free-trader Ronald Reagan became President in January 1981, the first thing he did was to limit the ability of the Japanese to export cars to the U.S. (a move that would soon be replicated in Canada and Europe).
We, however, had seen enough by that time to be certain that there was something dramatically different in the way Toyota and other Japanese companies worked with regards to product development, production, supplier management and general management. We just didn't have the evidence to make the case that we wanted to make. So, we started our research in the International Motor Vehicle Program in 1985 to make performance differences clear and trace them back to practice.
What put us in a different league from everybody else was that we were convinced from the beginning that it was a system issue, not an issue only at the plant level. Dan Roos, Dan Jones, and I were not traditional managers and therefore had the flexibility of mind to see the interactions happening between departments at Toyota and conclude that the key to its superior performance was not just a specific set of practices on the shop floor but an enterprise-wide philosophy and new methods in every element of the business. That was the real contribution of Machine.
RP: What about the community’s ability – or willingness – to carry on the legacy of the book and spread Lean Thinking?
JW: In Corporate Culture, Edgar Schein argued that there are three elements to organizational culture: artifacts (or tools), expressed values, and basic assumptions (or beliefs). It’s fairly easy to copy tools and write down the principles behind those tools, but when it comes to a system of beliefs, things get more complicated.
We saw this after Machine came out. People dashed off to find tools that could help them (like the ones described by Richard Schonberger in Japanese Manufacturing Techniques), often supported by consultants. And the companies’ mission statements were already full of principles. Organizations then set out to apply the lean tools using corporate programs that stated their guiding principles and that were led by continuous improvement teams. This approach turned out to be pretty ineffective. It didn’t shake up the belief system in place at most Western companies, whose focus was almost invariably on asset utilization and uptime and top-down control – both in clear contradiction to Lean Thinking.
I remember visiting a Cadillac plant at the end of the 1970s. The plant manager was very proud of the factory’s performance, which was captured in his most important metric – uptime. He didn’t seem to care much about the huge number of defective vehicles sitting at the end of the assembly line. When I pointed out to him that his assembly plant was busily making mistakes to be corrected by the nearby re-work building run by the sales organization, he simply noted that the number of defects wasn’t his metric. While to lean thinkers this contradiction was obvious, it just wasn’t (it often still isn’t) in many people’s system of beliefs. That’s why transferring ideas that seem pretty simple is actually really hard!
RP: Back to the book, was there one part of it that was harder for you to write? Were there things you were observing that you found harder to analyze and understand at first?
JW: One part that was really hard to write the chapter on dealing with customers. It dawned on us very early that there were very different approaches at Toyota and at Western carmakers, but the problem was that there were no data we could look at.
We saw Toyota’s efforts through its sales organization to level demand – rather than create an end-of-quarter surge – and their way of using the selling process to learn something about their customer as markedly different from Western firms. It was impressive how they expected to have customers for life and periodically sat with them to understand their needs. We were also struck by how Toyota forged life-long partnerships with the big dealer organizations they entered joint ventures with. We felt this chapter was very, very important, but didn’t have any data on performance in car dealing. We got to early 1990 with nothing and our deadline looming, and I remember Dan going on a last-minute trip to Japan to get the information we needed for that chapter. (He later turned what he learned into a remarkable program on international car distribution and laid the basis for the equally remarkable work Dave Brunt and John Kiff did at the Lean Enteprise Academy on creating lean dealers.)
The other thing we couldn’t write about was how management works at Toyota, which later proved to be central to the development of a lean enterprise. We just didn't know anything about it. Going to look at transplants didn’t tell us much either, because the feedback Western managers got from their Japanese “shadows” was never written down or at least it was never available to the public. So, we didn’t know how to write about it and, to be fair, we didn’t understand the importance of writing about it. We would have saved everybody a lot of trouble if we had gotten early on into more detail about lean management as the alternative to modern management (GM) and traditional management (Henry Ford)!
RP: Lean brought a revolution to the automotive industry and, later on, to many other sectors but with varying different degrees of success. What are your thoughts on how it spread to different types of activities?
JW: Machine doesn’t talk about other industries, but we implied that the principles described would work anywhere. So, it’s not surprising that, when the recession of 1991 hit, a lot of people in a lot of industries picked up the book to get inspiration on how to do things better.
At that point, Dan and I decided that rather than doing a sequel to Machine, we would go out and set up missionary societies for the explicit purpose of transferring these ideas to every value-creating activity. This probably contributed to the spread of Lean Thinking to other sectors. As these ideas were transplanted, however, they were adopted with different levels of accuracy and usually only narrowly in production, which led to different results.
Construction is an interesting example because they were really the first industry out the gate. The single biggest problem there was that most of the work was done by independent subcontractors who were also working for all the competitors of the general contractor. This made it hard to do scheduling, as subcontractors were diverted elsewhere, materials did not arrive on time, and customers changed their minds about was actually wanted. So, lean thinkers came up with the Last Planner System to try and bring visibility each morning to what was supposed to happen that day at each job site. (These were a curious contrast to the daily management meetings that soon emerged in manufacturing to review what had gone wrong the previous day and prevent it from happening again.) Last Planner was a brilliant effort to damp chaos by having all of the subs stand face-to-face daily with the general contractor and commit to a piece of work for that day. But continuing chaos in construction more generally has made it difficult for other lean methods and beliefs to take root.
Then there is healthcare, where the very high cost of running hospitals has made the focus on asset utilization (requiring massive amounts of queuing) a tremendous hindrance to lean transformations. There is also pervasive traditional management thinking, with senior doctors often struggling to think collaboratively (with many instances of “do as I say” rather than “let’s analyze this together”.) There have been enormous numbers of lean experiments in healthcare. Many of them have been hugely successful when they put the quality of care and the experience of the staff at the center of the improvement work. But they have proved difficult to sustain without looking at healthcare as a complete lean enterprise with fundamental beliefs aligned with stated principles and lean tools.
These are just two examples, and the translation of lean ideas from manufacturing was not easy in either case. In both instances, we have seen that building the management system and beliefs that can support the lean tools and principles is not a trivial problem.
I haven’t mentioned lean in government, logistics, services of many types, and so on, but the story is the same: taking a set of ideas from one industry and transferring them to another is never a slam dunk. It always requires adaptation and experimentation. And most experiments – in organizations whose principles endorse risk-taking but where everyone actually believes that experiments must never fail – don't work very well the first time out. So, progress is slower than it could be.
RP: What about the digital industry? Where does it stand in terms of lean adoption?
JW: There has been a lot of reluctance in the software industry to embrace lean language – even though methods like Scrum and Agile are conceptually lean. Part of the reluctance stems from the fact that the software industry sees itself as the most sophisticated, modern industry there is – a far cry from the old-school, polluting automobile and it factories. And part of the problem is that the lean community has not tried very hard to engage the software/digital community. It’s been busy elsewhere. And that’s fine. My interest is in making the world a better place, not making “Lean Thinking” a stronger brand, and I think the spread of lean ideas across the digital economy is proceeding more widely and successfully than most observers imagine.
At the same time, it is clear that digital has completely disrupted a lot of traditional industries, including automotive. Cars are now primarily repositories for massive amounts of code, which is why legacy carmakers have to make a leap. And the world increasingly looks for mobility as a service rather than cars an object, which requires another leap.
Tesla (not coincidentally based in Silicon Valley) has certainly been challenging traditional assumptions in the car industry by stressing connectivity, low-carbon energy, autonomy, and asset sharing rather than manufacturing efficiency and low levels of delivered defects. (Just as Tesla passed Toyota as the car company with the largest share value, it finished dead last in the J.D. Power quality survey that we used extensively in Machine to show the superiority of Toyota compared with American and European competitors.)
Toyota is very much aware that the world has changed and is working hard to figure out how to apply its thinking to the new mobility rather than the old automobility. This involves concurrent engineering in activities as disparate as code writing (using TPS principles for the massive software suites needed for autonomy) and in alternative motive power for vehicles (hybrid, full electric, fuel cell). All of us in the Lean Community need to pay attention and to engage in our own experiments as we can.
RP: What is the role that Lean Thinking has to play in such a complex world?
JW: Is it true that the world today is more complex than before? I’m always skeptical about claims that we live in exceptional times. But the world certainly is complex, and what makes lean so powerful is its ability to provide clarity about the flow of value through complexity and to stabilize the flow in preparation for improving it. That’s the objective of much of what we do. Indeed, it is the job of most management to stabilize the flow of value, even in the presence of a few managers who love chaos (from Henry Ford to Elon Musk) and can pivot quickly in times of uncertainty.
There might be situations in which it is tempting to have a chaos-loving, dictator-style managers come in for revolutionary change. But I believe that in the long term, the companies that can create stability out of chaos are going to be better off. Most of the world, most of the time, is either pretty stable or could be stable if only people would just make it their business to stabilize it. That’s what Lean Thinking can help us do: build a platform of stability (involving daily management and A3 analysis) that actually makes us flexible and ready to adapt to change (through hoshin planning) when it’s needed.
Lean is often perceived as a slow idea, especially in a world where everybody wants fast solutions and miracle cures – most recently, vaccines. But are we really that flat-footed and slow? I don’t think we are, and I think that more people will appreciate that soon. At this moment, the world might seem more interested in companies like Tesla than in companies like Toyota (and it’s important we stay humble and understand why). But we must stay true to our beliefs – that sustainable change doesn’t happen instantly and through individual heroic actions alone, but instead through collective efforts to countermeasure problems on a continuing basis.