THE BOTTOM LINE: Debate within the lean community often focuses on how fast or slow to tackle problems and undertake change. Art Smalley, LEI faculty member, Toyota veteran, and author, argues that fast or slow misses the point: Within an organization, different levels of problem solving exist — from near-instant responses to problems on the frontline to leadership planning for and pursuing a new business objective — and the cadence at each level is not arbitrary but determined by the type of work, mechanisms in place that support and coordinate problem solving, and the type of problem being addressed. What matters is that better thinking produces faster quality of results, not speed for its own sake, but speed that comes from reducing learning friction, confusion, and rework at any level. The quality of the outcome reflects the quality of the thinking behind it.
There is a well-known corporate slogan inside Toyota: よい品、よい考え (yoi shina, yoi kangae, or roughly translated “good thinking, good products”). It is a simple phrase, but it carries a profound operating principle. The quality of the product reflects the quality of the thinking behind it. A close corollary, and one that I think sometimes gets lost in discussions of the Toyota Production System (TPS) and lean thinking, is that better thinking also produces faster quality of results. Not speed for its own sake, but speed that comes from reducing the friction, confusion, and rework that slow organizations down in the first place. This point gets obscured when the debate turns into whether we should “go fast” or “slow down.” Both camps miss what Toyota actually teaches. The real question is: what mechanisms enable better thinking to occur?
I want to explore this through two experiences from my own career. One is a national laboratory where I have consulted for almost two decades, an environment filled with brilliant scientists where the culture of deliberation did not produce the results one might have expected. The other is Kamigo Engine Plant in Toyota City, where I worked over 37 years ago, and where I first learned how standards and documentation also serve as a hidden enabler behind both speed and quality. The settings could not be more different, but the lesson is the same.
The National Laboratory Paradox
Quite a few years ago, I was asked to help a national laboratory reduce the lead time for a complex, high-tech product vital to national security. Consistent timely delivery was a problem, and it generated a lack of trust between various oversight entities. The
organization was filled with brilliant people, physicists, engineers, and technicians who were among the best in their fields. Nobody was rushing them. The culture valued deliberation. “Slow down and be careful” was the ingrained institutional reflex.
And yet the lead time to produce this product was long, highly variable, and inconsistent. Yields on complex processes varied, equipment was not available when needed, and only certain experts could operate certain processes. End customers in the military defense supply chain and high-level administrators at the Department of Energy were often not happy. Internal management was not happy either. Something wasn’t working.
The natural assumption might be that the lab needed to speed up. But that wasn’t the real problem. Speed measures rate but it does not measure quality. Smart people working carefully and independently were still producing inconsistent outcomes, struggling to coordinate, and repeating problems that should have been solved once and stayed solved. The old TPS depiction with two pillars with both a time component (JIT) and a quality of condition component (jidoka) exists for this reason. Relying upon one alone is not sufficient in most environments.
What was missing wasn’t intelligence or patience. Or even a speedometer governing work cycles. It was a combination of mechanisms, specific standards, management structures, and routines that make thinking reliable, transferable, and improvable.
The Misconception About Speed
In the lean community, I often hear two opposing voices on this topic. One camp emphasizes speed: short cycles, rapid experiments, bias toward action. The other emphasizes slowing down: reflect more carefully, don’t rush to judgment, go deeper before you act. Usually these approaches are artifacts of the individual’s own personal learning style. Both instincts contain some truth and are compatible with TPS and lean thinking. But taken alone, both tend to miss a deeper point. The Toyota corporate slogan is “good thinking, good products,” and not simply faster thinking or slower thinking equals good products.
People can make mistakes at high speed. They can also make mistakes at low speed. A team can move slowly for months and still arrive at the wrong conclusion. A person can rush through a problem and get lucky with the right solution. Time by itself is not a countermeasure. Slowness is not a management system.
People can make mistakes at high speed. They can also make mistakes at low speed. Time by itself is not a countermeasure.
The real question, the one that I think gets obscured in these debates, is more fundamental: what enables better thinking and actions to occur in the first place?
The Toyota answer for decades has been to build standards, routines, and mechanisms that improve the quality of thinking in actual work that leads to actual performance results. When those mechanisms are strong, people can respond quickly when speed is required and think more carefully when the nature of the problem demands it. That distinction matters. And over the past two decades the national laboratory experience made it very concrete for me.
What the Lab Had to Build
The fix at the laboratory was not a single initiative. It was building mechanisms at different levels, each operating at a different cadence.
At the shop-floor level, like most organizations, multiple basics were missing. Visual controls were weak or absent. Equipment operations and tooling standards were underdeveloped. Maintenance and calibration procedures lacked the rigor needed for consistent execution. Training for operators was informal and uneven, nothing close to the structured approach of TWI Job Instruction. There was no andon-type system for surfacing problems in production in real time. Without these foundations, the people closest to the work had limited ability to detect abnormalities, respond consistently, or prevent recurrence. Without the fundamentals to undertake what I call “Type 1” problem solving (reactive problem solving based on quick responses to immediate symptoms), it was in some ways a “worst of worst” situation with slower reactions and low quality of outcome.
In the middle layer, where the scientists and engineers lived, the gap was different but equally consequential. These were extraordinarily capable people. But when a problem arose, each person defaulted to their own preferred approach. One might design an elaborate experiment. Another might analyze data in isolation for weeks. A third might convene a meeting and talk through hypotheses without a structured method for narrowing down to a countermeasure.
The result was predictable: slow progress, poor communication across groups, and problems that lingered far longer than they should have. What was needed was not less thinking or slower thinking but a common structure for better thinking, specifically a standard problem-solving methodology for what I term “Type 2” problems (known problem, unknown cause), supported by A3 thinking as a communication and management discipline. A3 reports didn’t constrain anyone’s intellect. They gave brilliant people a shared structure so their insights could connect, build on each other, and lead to coordinated action. That structure mattered enormously.
At the strategic level, the lab also needed to improve its annual management cascade of improvement themes aligned to its mission. Not everything could be worked on simultaneously. As I often like to point out, there are infinite problems to tackle and finite resources. The best organizations prioritize wisely where they spend their time and energy. Leadership had to set direction, prioritize, and create a rhythm of review that connected the shop floor and the engineering middle to the organization’s purpose.
In other words, the lab learned to manage at multiple speeds: daily mechanisms on the floor, weekly and monthly problem-solving disciplines in the middle, and annual strategic alignment at the top. Each level had its own cadence. Each cadence required its own mechanisms. None of them worked by simply telling people to be more careful.
Standards at Toyota for Faster Thinking
This situation harkened back to my early days with Toyota in Japan in engine manufacturing. Kamigo Engine Plant was for many years the model plant for TPS inside Toyota. Taiichi Ohno was the founding plant manager. Most managers had roots to him and not just ceremonial handshakes, etc., as is often the case. Kamigo had primary materials, precision machining, assembly, testing, tightly coupled supplier delivery systems, etc. The engine and transmission are the most complex systems in a vehicle and require extreme rigor and attention to detail.
When I was helping support the startup of the Georgetown engine plant in Kentucky, we created approximately 20 different types of equipment documentation standards to transfer overseas. Keep in mind that equipment in machining is automated. You can’t manually make a crankshaft. A typical production line has over 300 machining steps involved, and today’s quality tolerances are just a couple of microns. This exercise was not a binder of generic procedures or mere human standardized work. It was a specific knowledge architecture, covering tooling, maintenance, equipment operation, hydraulics, lubrication, programmable logic control (PLC) circuits, computer numerical control (CNC) programs, spare parts, and troubleshooting. The goal of all this documentation was for detailed machine-level problems to be recognized, communicated, and solved rapidly and consistently across the ocean. I still have a copy of the submission specifications we got officially codified into Toyota’s Machine Tool Standards (MTS) and Toyota Manufacturing Standards (TMS).
Without that documentation, a person overseas facing a machine problem could stare at the equipment, rely on personal memory, ask around, and experiment through trial and error. Hours could pass. The same problem might get solved differently by different people on different shifts, with no learning captured. This is where both fast and slow thinking on their own have limited value. Time is merely a reflection of the efficiency of the actual work performed.
With high-quality documentation and standards in place, the same person could compare actual conditions to known conditions, narrow the possibilities, and often solve the problem in minutes. Not because they were rushed, but because the detailed standards mechanism reduced learning friction and supported better thinking. The first thing we did when we had a precision problem in machining was to compare actual conditions to our documented standards, such as cutting conditions, tooling conditions, material removal standards, operating conditions, machine static accuracy (e.g., spindle bearing run out). When we were at “standard” the machine would run to specification. With age, degradation, damage, or incorrect operations, however, there are always going to be problems to solve.
That insight is a key difference between standards as bureaucracy, which is how many people outside Toyota misunderstand them, and standards as stored learning. An excellent standard captures what experienced people know and makes it available to others. It creates a basis for comparison. A “gap-from-standard” condition creates a basis for action. Reflection on that action in turn creates the basis for a stronger future standard. That is why we realized we needed better equipment documentation for Georgetown than we relied upon in Japan. Experienced workers with tacit expert knowledge did just fine in Japan. For Toyota to take complex precision machining operations overseas, however, that tacit knowledge base needed to become more explicit in the form of a clear standard. We rewrote the standards based upon the requirements of the actual situation.
Standards as bureaucracy is how many people outside Toyota misunderstand them. Standards are stored learning.
This codification of knowledge is part of the hidden engine behind Toyota’s speed. It is not that people move recklessly fast. Conversely, they don’t sit around pontificating based upon idle conjecture. Toyota has found in design and development, manufacturing, assembly, and testing that when the right mechanisms and thinking patterns are in place, better thinking can happen faster with less friction.
Different Gears, Different Mechanisms
Another way to think about this is to realize that by design of the management system Toyota does not operate at one speed. An element of work on the assembly line may be measured in seconds. Takt time drives the rhythm of production. Daily huddles and shift reviews operate at a different cadence. Weekly and monthly planning meetings operate at another level. New product development may stretch across two to four years.
These are intentionally different gearing ratios. And the cadence at each level is not arbitrary. It follows from the type of work, the type of governing mechanism used for coordination, and the type of problem being addressed.
I have written elsewhere about four types of problems:
- Restoring operations quickly to standard,
- Diagnosing an unknown root cause and solving it methodically,
- Improving toward a target condition both rapidly and slowly, and
- Open-ended innovation.
Each type involves different tools, modes of thinking, and a different tempo. Restoring a process to standard needs to happen quickly and, hence, is governed by an andon system that requires speed in response. Conversely, there is no immediate andon type system for the top level of
hoshin kanri or product development; different tools, forms, meetings, and thought patterns signal problems and govern those situations, like an obeya room, for example.
In either case, the pace is not the key point. The mechanism and thinking pattern are the points. The right mechanism at the right level enables the right quality of thinking at the cadence the situation requires.
The Point for Practitioners
When I reflect on decades of work, from Toyota engine plants to national laboratories to organizations around the world, the pattern is consistent. The organizations that think best are not necessarily the fastest or the slowest. They are the ones that have invested in the mechanisms that make thinking reliable: standards, visual management, structured problem solving, defined escalation, regular review cadences, and learning routines grounded in PDCA cycles for improvement. PDCA can be fast and it can be slow.
The right mechanism at the right level enables the right quality of thinking at the cadence the situation requires.
Where those mechanisms are strong, people can move quickly with confidence. Where they are absent, even the most talented people struggle, no matter how much time they are given. So when someone tells you the answer is to speed up, ask: speed up with what mechanism, and to what standard of performance? And when someone tells you to slow down and think more carefully, ask: think more carefully using what structure, and how exactly does that relate to the quality of the result?
The answer is rarely simply faster or slower. The answer is better thinking, supported by the standards and mechanisms that make it possible. Or as the Toyota corporate slogan implies “yoi shina, yoi kangae” — “good thinking, good products.”
Not speed alone. Not slowness alone. But better thinking that can happen faster, with less friction, in service of the work that matters for performance.
Building a Lean Operating and Management System
Gain the in-depth understanding of lean principles, thinking, and practices.
“Good Thinking, Good Products” – Toyota’s old slogan still wins.
Structure is the real accelerator indeed, Art! Because, speed without structure is just expensive chaos.
As you correctly reminds us… better thinking isn’t about going faster or slower. It’s about having the right mechanisms and standards so good thinking happens reliably, every time. Once you install those mechanisms that turn good thinking into a repeatable system, real impact follows naturally.
Keep on posting!