Toyota's real secret sauce
FEATURE – The hardest part of a learning journey is learning to learn and figuring out what we need to learn - as opposed to we want to. The author wonders whether we are drawing the right lessons from TPS and highlights a few things we are underemphasizing.
Words: Michael Ballé
In 2022, Toyota sold 1.12 million Corollas and 0.87 million Rav 4. In comparison, the Ford F-series sold 0.79 million and the Tesla model Y 0.76. In total, Toyota sold 10.5 million vehicles last year, far ahead of second-rank rival Volkswagen with its 8.3. And yet, Toyota is in trouble – again. Actually, Toyota is always in trouble. It botched the full electric battle and is losing its leading edge. It is natural to ask, therefore, whether Toyota has met the fate that any dominant company eventually meets and fallen to its younger rival Tesla because of hubris. Maybe. We shall see. For now, the Corolla is still the undisputed best-selling car in the world.
For those of us who believe that lean is the Toyota Production System (TPS) adopted out of Toyota, we need to regularly ask ourselves two hard questions: first, is TPS still the best business model available? Second, do we draw the right lessons from what we see at Toyota and understand TPS to be?
TPS might no be the best business model forever – maybe the “Tesla way” will beat it in the end. But TPS is so far the only framework that can be studied and experimented with, and the only viable alternative to finance (top line appears by magic, operational costs must be reduced by squeezing budgets, financial activities is where the real profit is and anything unexpected can be counted as exceptional costs).
TPS is a learning system that tells us to constantly question:
- Can we bring more value to customers?
- Can we stop defects closer to where they are created?
- Can we reduce lead-times between orders and deliveries?
- Can we better balance workload and capacity?
- Can we engage people in studying their standards and try small steps to do things better?
- Can we further develop mutual trust with employees and suppliers by joint problem solving and making sure systems work as they should?
Because these core questions build on each other, asking them demonstrates that TPS is indeed a learning system. For instance, I would not recommend trying to reduce lead-times without first looking into self-quality, and vice-versa, working on quality responses is hard to do without the pressure of just-in-time.
The TPS has been written about extensively and is very well studied – but how does it relate to the Corolla being the world’s best-selling car year on year? What is it that we see Toyota do that we don’t fully integrate in our thinking? Reflecting on the past decade of lean experiments and debates, I can see a few points we tend to under-emphasize. I believe that in them lies the answer to the above questions.
VALUE IS CUMULATIVE
First, it’s all about products and customers. The aim of any business framework is to make better products that can be sold longer and more profitably. It’s very easy to get distracted by all sorts of internal operational issues and miss the elephant in the room (actually, work around it): products or services must beat the competition in the customer’s minds, hearts, and wallets. Toyota’s approach to this is unique. It doesn’t look for the brilliant products that will take over the market and swoop up all the chips. It carefully – painstakingly, one could think – generates value for customers over several generations of products.
Top Gear’s car maven Jeremy Clarkson has been delighting us with great videos about how he hates the Corolla, the best-selling car ever because it is so… boring (“as interesting as the night sky when it’s cloudy” is one of my favorites). Best-selling worldwide since 1997? How can that be? Well, if we forget organization for a moment and look at product, we can see that the TPS has a definite framework for designing them: what value must we create? What waste should we eliminate? Rather than start with a blank page, they constantly ask themselves what extra value they should bring (considering current customer tastes) and what pain points for customers they could take away (including price). As I write those lines, I wonder: what am I saying that adds to what’s been previously written about lean? What waste should I eliminate? Should I really recount the steps of TPS every time? Clarkson’s point is that the Corolla sells not because it’s the most exciting car on the road, but because Toyota systematically took out the pain points from the Corolla driving experience, and most drivers seek ease and safety rather than getting a rush from the road – as well as reliability and residual value.
This means starting with a full understanding of previous models as well as competitor moves. It’s not perfect. One explanation for Toyota’s fiasco with their full electric car is that they tried to make it fit in the hybrid platform assuming the solid-state battery they are working on would work (it does, but its autonomy is poor). The TPS brings them back to square one: tear down Tesla’s bestselling models. Will Toyota engineers draw the right lesson from going back to basics? Time will tell, but the method is clear.
DISCIPLINED MAINTENANCE, PRUDENT INNOVATION
Products and services are made and delivered through processes. Processes require both working systems and trained personnel to work as they should. Processes misalign as tools get blunted, parts get used, dirt gets in, new materials don’t react as expected, etc. People then get confused, lose focus, don’t know what they don’t know, forget what they should know, have good days and bad days, and change their minds about what they’re doing and why. The work environment changes daily and presents new challenges every other day. As a result, nothing ever goes quite as planned.
A process is a sequence of dependent events that go exactly as planned. This, in real life, is a very rare case of how systems behave. In a system, components react to each other and rarely exactly as predicted. The more one puts pressure on an element of the system, the greater the response of the others – often not at all what was anticipated. A 20-year obsession with processes has left us blind to systems, which behave more like living ecosystems than machinery. The only concrete way to keep a system in control is to perform maintenance every day, right away: both checking that things work as they should and correcting abnormal situations right away. Doing this, you both keep the system under control as well as learn about it and how it behaves.
Most Toyota plants in the world produce on two shifts and do maintenance during the night shift. This makes total sense, as a night shift is necessarily more expensive than a day shift and we all know that night shifts are a world of their own. However, we often fail to draw the deeper lessons from this practice: overinvesting in maintenance, in order to create performance from working systems, and trusting in these systems because you know them.
The very first presentation I came across on kaizen by Professor Masaaki Imaï in the late 1980s introduced kaizen as the frontier between maintenance and innovation.
The current societal obsession with digital innovation makes us forget how deep this chart really is. As a mental experiment, imagine our society if we had doubled the investment in maintenance in hospitals, schools, trains, or R&D for the past 20 years?
THERE IS ALWAYS TIME FOR A TEACHABLE MOMENT
If you exclude die-hard taylorists, who see in the lean tools ways to perfect processes by imposing more detailed work rules on people (and there still are many), most serious students of lean have accepted that “to make products, first we need to make people” is the bedrock of Toyota’s paradigm change. Yet, it’s not easy to see what this means in practice. It’s hard to know where Toyota is at now that it has become such a large company. Its veterans, however, all remember the teachable moments when their boss would stop work in order to show something, discuss it, demand kaizen, or just challenge a situation.
This is not as simple or easy as it sounds. A few days ago, I was in a banal steering committee and a point came up that needed clarification and discussion as it was central to the mission of the organization – something everyone should understand and have an opinion on. It wasn’t an easy discussion because it wasn’t an easy topic, but it was very much a learning conversation. Still, the pression to return to the agenda and get through micromanagement decisions was palpable. Prioritizing teaching moments over production requires real commitment.
First, it assumes that the senior person has something to teach and knows how to do so – very few jobs are thought that way. MBA programs teach people to optimize current processes and budgets in order to better exploit narrow activities now, rarely to prepare teams for tomorrow. Secondly, people have to be open to instruction. For this, we need to think deeply about how the work environment makes room for readiness, encouragement of learning and deliberate practice (the three basic laws of learning as identified by early researchers) – or not.
Learning and teaching as a key part of any job is a core component of the lean worldview and requires some working theory of how to both learn and teach. The dominance of the give instructions/check execution theory of command-and-control doesn’t leave much room for learning or teaching time, nor space to understand that connecting the dots rarely happens when someone is focused on doing a job but, on the contrary, when they are interrupted by a new idea, either from another domain or from putting two and two together differently on routine tasks – the whole point of kaizen.
ALWAYS ASK: WHAT IS TRUE QUALITY?
“What is a good app?” “What is good service at a hotel?” “What is a good espresso machine?” “What is a good plastic molded part?” Quality is always taken for granted because quality is so hard to define. It is, however, what makes sense to customers and what customers ultimately look for and buy (it can also mean different things to different customers or even to the same customer at different times or with different uses). Because the “quality question” is so open, we tend to close it by defining “standards.” With these standards in place, the question shifts from “are we doing the right thing” to “are we doing things right”. Clearly, in an ideal world, we want to do the right thing well.
Because doing things well is easier to spot compared to doing things poorly, managers tend to focus on solving that question and no longer wonder about doing the right thing or doing the wrong thing. All the more so that, usually, doing the right thing is hard, and hence at first it will be done poorly until learning kicks in.
The issue is that true quality lies within doing the right thing, not doing the wrong thing well. The true quality question is never closed: challenge product standards, challenge material and parts standards, challenge education and training standards, challenge tolerances, challenge instrument settings. Do these standards actually deliver quality for customers? Or do they serve to reassure the organization that all is well? The aim of challenging standards is to investigate measuring methods and question the relative impact of quality characteristics against customers’ opinion of the product until you reach a consensus on defects and flaws of the current production process.
The aim of asking about “true quality” is to expose latent defectives, pieces of work that we think are ok but in practice won’t do the job for the customer. To do so, we need to visualize the customers’ own processes, see where, when, and how they use our work and identify what helps them and hinders them. In other words, where the pain points are. Brace yourself: this will always be a difficult discussion, simply because it is such an open question, and there are more mysteries than facts. Still, one of the key secrets to making winning products is to constantly question quality standards and discuss endlessly what, exactly, is true quality?
REVEAL REAL PROBLEMS TO DEAL WITH THE DIFFICULTY
At the heart of any problem is a true difficulty: physics won’t allow a process, machines won’t have that kind of precision, someone won’t do or doesn’t know how to do what they’re expected to and so on. Rather than stop at this difficulty, the aim of problem solving is to surface it and then find a way to use it to do something else. To be creative. To come together and look at the situation differently and find a way around the obstacle. As Duncker’s example visualizes, if the main road to the fortress is too well defended, find another way in. Use the difficulty to discover what can be done differently.
Productive discussions tend to be difficult ones. Since Adam Smith’s pin factory in the late 18th century, organizing has been about dividing labor to benefit from specialization and then avoiding duplication of labor, to seek economies of scale from volume. This arrangement makes sense in stable situations where volume is guaranteed; but it is also terribly inflexible and produces heaps of waste in terms of rework, inventory, backlogs, overinvestment, and endless problems. Yet, it is also useful to avoid conflict by separating activities and keeping them to different teams, departments, or functions so that everyone can go on producing without ever considering that, to take Kaoru Ishikawa’s phrase, the next process is the customer and wondering about what true quality means for the proximal customer as well as the final one.
Organizations avoid facing environmental changes or internal challenges through various defensive routines: they focus on side issues, in order to distract from the real “elephant in the room” issues and to avoid authentic conversations about doing the right/wrong things. These defensive routines allow the organization to continue to operate, whether it performs for customers or not. It’s not that organizations don’t confront problems, more that they choose problems to which they have workable solutions to – what used to be called “garbage can theory” (an organization is a collection of choices looking for problems, issues and situations in which they might be aired and solutions looking for problems they might answer).
Lean techniques such as kanban, jidoka, or standardized work are all about revealing the real problem, the gap with ideal delivery to set the scene for genchi genbutsu: go to the site, see the real problem, listen to the people in their context, create consensus on the real problem and obtain a commitment to kaizen. True kaizen is bringing value closer to the customer, which is hard because things are how they are for a reason, and so we must accept that for the team the difficulty is the way, the difficulty is the opportunity to understand in greater depth and find an alternative. Kaizen bridges de gap between maintenance and innovation.
SPLIT THE WINS AND BUILD TRUST
When I first studied Toyota’s approach to engaging with suppliers, I saw Toyota engineers going on a monthly visit to a supplier’s line to help it improve quality, delivery, and productivity (it makes perfect sense in terms of maintenance and kaizen). Over two years of continuous improvement, the line boosted its productivity by 30% and reached flawless delivery through high flexibility (batches of less than one hour). Toyota never asked for any of this productivity improvement in price negotiations, and I wondered where the gain was other than securing supply, which, agreed, has a much higher value in the TPS than in other production systems. It turned out, however, that the work on the line led engineers to redesign the product and process for the next generation product so they could achieve a total cost reduction of 30%! Toyota then split the difference, keeping 15% and leaving 15% to the supplier. Imagine doing this product cycle after product cycle?
This was extraordinary in two ways. First, Toyota engineers knew where to look for the gain, and it wasn’t the direct productivity of the line but what could be learned from it for the next line. Then they knew to split the win to build trust to do this again on the next cycle. Once I realized trust was a fundamental element of the system, I started seeing it everywhere in the way Toyota operates. At one point, the supplier no longer shipped parts to Toyota. Still, the Toyota sensei went to visit the line and discuss with his deshi at the supplier. “Why would I not?” he asked puzzled when questioned about it. Good relationships are more important than day-to-day business.
Trust is the juice that runs the human engine. You want to associate with people, whose competence you trust because you trust them to do the right thing and win. You want to associate with people you can trust to share their wins with you, who care about your circumstances and what happens to you. Building relationships means both finding gains and sharing them.
Trust is not a “nice to have” in the human experience. It is central to it. Trust that people have your back, trust that machines work reliably, trust that you understand how the world works and trust in your own focus to build the work of art that is your life. The whole TPS system is a learning system to build trust, with customers, with employees, with machines and sites, with suppliers, and so to ensure easier collaboration and flow, both in the work and in the psychological sense of intense engagement, focus and contentment in the present activity and the current moment.
IN AN EQUATION?
What has 30 years of discovery and experience with the TPS taught us, then? What should we train ourselves to deliberately focus on every day in order to enrich our professional lives? If I had to sum it up in a pithy equation, I’d always keep these four components in mind:
- Where is true quality for customers and how to improve it?
- What is the path to improving true quality, with whom and how do we learn together?
- What are the wasteful activities that slow us down and we should reduce?
- How long will it take to do any of this, and can we accelerate?
Or something like:
I realize this takes us far from improving processes by making parts move faster in the material flow, far from the mechanical view of organizations, but isn’t that the whole point of learning journeys? Sure, we learn what we set out to learn, but how do we discover what we really need to learn? How to take the step back to see that what we find is limited by the questions we have and our narrow focus on what we came to seek in the first place?
What has Toyota really added to our knowledge on how to run organizations effectively, sustainably and responsibly? This is a harder question than one might think at first glance. I once had the opportunity to ask Mr. Takehiko Harada, who had worked directly with Taiichi Ohno, what he thought the West had gotten wrong about TPS. “Our first aim was training of subordinates,” he answered. “And we sought total optimization rather than partial optimization.” To learn, the student has to be ready to learn, which means we only hear the answers to which we have questions for. To truly learn, we have to seek the questions that fit the answers we’re offered.