Oscar Roche was joining live from Australia. It was the middle of the night for him. Jim Huntzinger of Lean Frontiers had been scheduled to co-present and was unable to join, so Oscar carried the session alone — which turned out to be fine because Oscar has spent more than a decade as a working practitioner of both Training Within Industry and Toyota Kata and was prepared to take the audience through how he sees them fitting together.
The framing Oscar opened with sat on a small handwritten sign next to his computer: "If you can't explain something simply, you don't understand it well enough." That challenge — to explain simply — was what he was bringing to the session. The premise underneath was that TWI and Toyota Kata aren't competing methodologies or even just complementary toolsets. They're skill patterns that work as countermeasures to specific symptoms that show up in any organization trying to build a culture of improvement. When you understand the symptoms, the choice of which skill to apply becomes obvious. When you don't, the practice becomes random, and random practice doesn't build culture.
The session is most valuable as a clear map of which skill addresses which problem, why PDCA sits underneath all of them, and how the coaching pattern at the heart of each of them is more similar than most practitioners realize.
Oscar Roche is Director of the Training Within Industry Institute and Visual Workplace Australasia, based in Australia. He holds a Bachelor of Applied Science in Dairy Science from Roxbury Agricultural College — a credential that led to a career in multiple chemical companies and food production firms. Oscar has worked as a manufacturing manager in many settings, and since 2010 has focused on work related to Lean, particularly Training Within Industry and visual management methods.
Oscar is a TWI Institute master trainer and works extensively with organizations adopting both TWI and Toyota Kata. He partners with Jim Huntzinger and the Lean Frontiers team on the annual TWI Summit and Toyota Kata Summit, where his TWI/Kata simulations have become a recurring feature. His framing of the work is grounded in practical delivery — what he's seen work, what hasn't, and what changed his thinking along the way.
Oscar opened with a definition that he uses with his clients: culture is the sum of people's habits as they go about their daily work. The definition is one among many possible definitions, but it's the one that ties most directly to the work of improvement. If culture is habits, and you want to change culture, you have to change habits. The easiest habits to change are your own. So building a culture of coaching starts with developing coaching habits in yourself.
The next question Oscar wanted to surface: when does a culture of coaching actually add value? His answer came through a reframe of the definition of Lean itself.
For a long time, including in Oscar's early Lean exposure in the mid-1990s at Ecolab and later at Madeira Fruit Juices, the working definition of Lean was the elimination of waste. The standard diagram of value-added and non-value-added work was what everyone learned. The problem, Oscar said, was that the improvements he and his colleagues made under that definition often didn't stick. Something was missing.
About four years before the webinar, Oscar encountered a different definition that changed his thinking: Lean is developing people who can and will solve problems and make improvements daily. The shift from "elimination of waste" to "developing people who can and will solve problems" reframes the entire endeavor. The waste-elimination definition makes the practitioner focus on the waste; the people-development definition makes the practitioner focus on the people, with the waste reduction following as a consequence rather than as the goal.
The reframe matches the discipline Toyota describes for its own improvement work: the first priority is developing people, and the second is the benefit of the improvement. Oscar's hypothesis, drawn from his own experience: organizations operating under the waste-elimination definition may get some value from TWI and Kata, but he's not sure how much. Organizations operating under the people-development definition find the skills snap into place because the skills are designed for that purpose.
Oscar's working language for who's involved in improvement work uses two roles. The doer practices the skill or habit that delivers the result. The coach helps the doer grow their capability as the doer learns through practice. The relationship can and should occur at every level of the organization. The improvement cycle is fastest at the front line — where the work happens — and slows as you move up the organization, where the improvement cycles at executive levels might be measured in months rather than minutes. The pattern of thinking and coaching, though, is the same at every level.
Oscar named an irony from his own work. The coach should ideally be the manager of the doer. The manager is the person closest to the doer's work, the person most invested in the doer's development, the person most likely to be present when coaching opportunities arise. In practice, Oscar said, he often finds himself or his team asked to be the coach — and finds himself thinking, in the middle of the session, that he isn't the right person for the role. The right person is the doer's manager. External coaches have a place, but their job is to help managers establish the framework, not to do the framework's coaching work themselves. The session returned to this distinction more than once.
Oscar's central framework is what makes the session distinctive. He doesn't present TWI's three skills or Toyota Kata's improvement kata as good practices to adopt because they're good. He presents each one as a countermeasure to a specific symptom that any organization will encounter at some point. The symptom diagnosis is what determines which skill to apply.
Symptom 1: "Don't know how, can't do." When a frontline worker doesn't know how to do a task and can't do it, the symptom is a training gap. The countermeasure is the skill of instructing — Job Instruction in TWI's framework. JI is a structured method for breaking down a task, identifying the key points, and teaching the work in a way that the learner can actually absorb and perform.
Symptom 2: "Don't care, won't do." When the worker has been trained, demonstrates that they know how and can do, but isn't doing — the issue isn't training. More training won't help, because the gap isn't knowledge. The symptom is an engagement, relationship, or leadership issue. The countermeasure is the skill of leading — Job Relations in TWI's framework. JR provides foundations for proactive leadership behavior and a four-step pattern for handling people problems when they arise.
Symptom 3: "Physical, hard to do." When the work itself is physically difficult, inefficiently organized, or harder than it needs to be, the symptom is a methods issue. The countermeasure is the skill of improving methods — Job Methods in TWI's framework. JM gives the frontline supervisor a structured way to look at how materials, manpower, and machines are organized and to improve the arrangement without waiting for an engineering redesign or a kaizen event.
Symptom 4: "Random or recreational stop-start improvement, not reaching goals." When the organization improves in ways that are unfocused, inconsistent, or detached from goals, the symptom is a thinking-discipline issue. The countermeasure is the skill of scientific thinking — the Improvement Kata in Mike Rother's framework. The Improvement Kata provides a structured pattern for setting direction, understanding current state, establishing target conditions, and experimenting toward them.
The framing matters operationally. When an organization comes to Oscar saying they want to "start practicing Toyota Kata" or "implement TWI" or "tie Kata to risk assessment" — which was a recent inquiry — Oscar's first question is why. Not as a rhetorical challenge, but as a genuine diagnostic. Why does the organization want to practice this skill? What symptom is it a countermeasure to? In 50 to 75% of cases, the response is a blank face. The discussion that follows that question is where the real value lives, because it surfaces whether the organization has identified a real symptom or whether they're chasing the methodology because methodology is in the air. If there's no symptom, the methodology will be a hard sell internally, people won't connect the dots, and the chances of sustained adoption are minimal.
The framing also avoids the failure mode where organizations try to "implement" a skill as a goal in itself. The skills exist to do something. They're tools that address specific problems. When the problem is clear, the skill makes sense. When the problem isn't clear, the skill becomes overhead.
The four skills aren't separate methodologies that happen to fit together. They share a common backbone: Plan-Do-Check-Act, the cycle Walter Shewhart developed and W. Edwards Deming popularized.
Job Instruction has a four-step method that maps directly to PDCA — prepare the worker, present the task, try-out performance, follow-up. Plan, Do, Check, Act.
Job Relations has a four-step pattern for handling a people problem that Oscar described as "PDCA for human beings" — get the facts, weigh and decide, take action, check results. Plan, Do, Check, Act.
Job Methods has a four-step pattern that's "absolutely PDCA-based" — break down the job, question every detail, develop the new method, apply the new method. Plan, Do, Check, Act.
The Improvement Kata has four steps that are also PDCA-based — understand the direction, grasp the current condition, establish the next target condition, experiment toward the target condition. Plan, Do, Check, Act.
The synergy isn't accidental. It's the predictable consequence of all four skills being designed around the same underlying scientific method. The diagram Oscar used to show the relationship — TWI's three skills as foundational and the Improvement Kata as a fourth — would more accurately fade one skill into the next rather than show hard boundaries, because in practice they overlap continuously. A given improvement moment might pull in Job Instruction (to train someone in a new method), then Job Relations (to manage the engagement of the team affected by the change), then Job Methods (to refine the standard once it's stable), with all of it operating inside an Improvement Kata cycle aimed at a target condition.
Oscar walked through each of the four skills with enough detail to make the structure clear. Job Instruction is the first.
The "get ready" steps for Job Instruction are four. Have a timetable or plan for training — who needs to be trained on what, by when. Break down the job into a recipe the trainer will follow — this is the most pivotal of the get-ready steps, because the job breakdown is what makes consistent training possible. Have everything ready — the materials, tools, and information the training will need. Arrange the worksite — the trainer always sets the primary example for how the workplace should look when the task is being performed.
The four-step method for the actual training session is where PDCA becomes visible. Step 1: prepare the worker. This step is essentially a Job Relations exercise inside Job Instruction — making the worker feel comfortable, building the relationship that the training will operate inside of. If the worker isn't comfortable, the training is going to be a tough road regardless of how well the trainer knows the task. Step 2: present the task. The instructor demonstrates and explains the task, including the key points. Step 3: try-out performance. The learner does the task back while the instructor observes and confirms understanding. Step 4: follow-up. The trainer checks in over time to confirm that the training is sticking and the standard is being followed.
Oscar quoted a line from David Meier's "Toyota Talent" that he treats as central: the identification and presentation of key points is the single most important aspect of the training process. Key points are identified during the job breakdown and presented during the training. Without clear key points, the training is generic. With them, the training is targeted at the specific things that determine quality, safety, and efficiency in this particular task.
The coaching pattern for Job Instruction — what a manager coaching a supervisor through their JI practice would do — follows the same PDCA shape. Is there a training plan? After the plan is set, what has the doer learned about the work and about the worker up to this point? When the doer delivers the actual training, what's the quality of the JI breakdown and the delivery? Based on the observation, what's the next step — more training, modify the breakdown, train a different person on the schedule, something else? And the question that Oscar emphasized appears in every coaching pattern: what do you expect to learn from that next step, and how quickly can we go and see what you've learned?
When Job Instruction is done well, the symptom of "don't know how, can't do" gets resolved. The workers know how and can do.
Job Relations addresses the next symptom — "don't care, won't do" — and Oscar was emphatic that his quick summary in a webinar context could not do justice to the depth of the skill. He described the week he spent in Train-the-Trainer Job Relations as the most valuable training experience of his life in terms of its impact on how he functions.
The skill has two parts. The four foundations are proactive — daily practices the leader follows to build and maintain the relationships that make the team's work possible. Let each worker know how he or she is doing. Give credit when due. Tell people in advance about changes that will affect them, and tell them why. Make best use of each person's ability.
The four-step pattern is reactive — what the leader does when a people problem actually arises. After a reorienting step, the pattern follows PDCA. Get the facts. Weigh and decide. Take action. Check results. Plan, Do, Check, Act, applied to a human situation rather than a technical one.
Oscar made a point about the relationship between the foundations and the four-step pattern that practitioners often miss. The earlier you start handling a problem — using the four-step pattern when something first surfaces, rather than waiting until it escalates — the more likely it is that the actions you take will be underpinned by one or more of the four foundations. The proactive practices and the reactive practices reinforce each other. The foundations create the conditions under which the four-step pattern can be applied effectively. The four-step pattern produces specific actions that often happen to embody one of the foundations.
The coaching pattern for Job Relations follows the same shape as the other skills. What's your most important people objective right now? What facts do you have? Based on what you've learned up to this point, what are your possible actions? What's your next step, what do you expect to learn from it, and how quickly can we go and see what you've learned?
When Job Relations is done well, the symptom of "don't care, won't do" gets resolved. The workers care and the workers do — what Oscar described as engaged.
Job Methods addresses the symptom "physical, hard to do." The skill is built around six questions that the supervisor asks about each step of an existing task.
The questions follow a particular order, and the order matters. The first question is why. Why are we doing this? What is the purpose of this step? If the answer is that there isn't a clear purpose, the step should be eliminated rather than improved. Improving a step that shouldn't exist is what Peter Drucker called "doing something really well that shouldn't be done at all" — Oscar's paraphrase of one of Drucker's lines that he treats as foundational to JM.
If the step is genuinely value-adding, the next questions are where, when, and who. Where should this step be done — is there a better location for it? When should it be done — could the timing be improved? Who is the best person to do it — does it require the skill level currently assigned to it, or could it be done by someone else more efficiently?
The how question comes last. How is the best way to do this step? Oscar emphasized that the structure of the questions is the point. Without the discipline of asking why, where, when, and who first, supervisors typically jump straight to how — and end up optimizing the execution of steps that shouldn't exist at all, or that would be much better done somewhere else, at a different time, by a different person.
The four-step pattern for Job Methods follows PDCA. Break down the job. Question every detail using the six questions. Develop the new method. Apply the new method.
The coaching pattern follows the same shape. What's your current plan with the use of JM, and why? What have you learned up to now? What's the quality of the job breakdown sheet on the current task? What's the quality of the proposal? What's your next step, what do you expect to learn from it, and how quickly can we go and see what you've learned?
When Job Methods is done well, the symptom of "physical, hard to do" gets resolved. The work is as easy as it's practical to make it given current resources.
The Improvement Kata addresses the symptom "random or recreational stop-start improvement, not reaching goals." The skill is what Oscar called the practice of scientific thinking.
The four steps of the Improvement Kata, drawn directly from Mike Rother's work. Step 1: understand the direction or challenge — get clear on where the organization is headed and what the long-term aspiration is. Oscar's image is binoculars looking at something in the distance. Step 2: grasp the current condition — examine the process closely and document how it actually operates today. Oscar's image is a magnifying glass looking at the work. Step 3: establish the next target condition — describe a specific, time-bound condition the process should reach by a defined date, as an interim state between current condition and the longer challenge. Step 4: experiment toward the target condition — run iterative PDCA cycles to address the obstacles between current condition and target condition.
The coaching pattern uses Mike Rother's five questions. What is the target condition? What is the actual condition now? What obstacles to the target condition are in the way, and which one are you addressing? What's your next step, and what do you expect to learn from it? How quickly can we go and see what you've learned?
The shape of the coaching cycle is the same as the other three skills. The fifth question — when can we go and see — is what creates the iterative rhythm. The coach and the learner aren't waiting weeks or months between checks. They're meeting daily, in short cycles, with each cycle informed by what the last cycle taught.
Oscar emphasized one specific point about the coach's role in Improvement Kata coaching. The coach has to ensure that the learner is connecting the learning from each step back to the next step. Without that connection, the experiments become random — the practitioner runs an experiment, learns something, and then designs the next experiment without using what was just learned. The connection between the learning and the next step is what takes the randomness out of improvement. It's what converts the work from experimentation-for-its-own-sake to goal-directed experimentation.
When the Improvement Kata is done well, the organization gets goal-directed experiments instead of random improvements. The system makes progress toward target conditions in a way that's visible, sustainable, and replicable.
One of the session's most useful framings came late. Oscar described how the four skills interact in practice.
In the Improvement Kata, obstacles surface as the team experiments toward the target condition. Some obstacles are technical — the process is physically hard to do, the methods aren't optimal, the materials don't flow well. The countermeasure for those obstacles is Job Methods. Some obstacles are knowledge-based — workers don't know how to do a task, or don't know how to do it the way the new standard requires. The countermeasure is Job Instruction. Some obstacles are relational or engagement-based — workers know how and can do but aren't doing, or the team isn't aligned around the work. The countermeasure is Job Relations.
The Improvement Kata becomes a pull system for the TWI skills. The organization doesn't practice Job Relations or Job Instruction or Job Methods because someone decided to implement them. The organization practices them when the experimenting pattern surfaces an obstacle that one of those skills addresses. The skills are pulled in by need rather than pushed in by management decree.
The same logic applies to other Lean practices — 5S, kanban, value stream mapping. The Improvement Kata creates the conditions under which those practices are pulled in as countermeasures to obstacles that surface in the experimentation. Without that pull, the practices have to be pushed, and pushed practices rarely sustain.
Oscar described an experience at a case study session in Allentown, Pennsylvania, where a presenter walked through a real improvement effort without realizing they were doing exactly this — surfacing a "don't care, won't do" symptom as an obstacle, applying one of the Job Relations foundations as the experiment, and resolving the obstacle. The presenter didn't have the framework Oscar was teaching. The pattern was visible anyway, because the pattern is how improvement work actually proceeds when it's done well. The framework makes the pattern teachable.
Oscar drew one specific parallel between Job Relations and the Improvement Kata that deserves its own attention.
In Job Relations training, the early sessions establish a model: an effective supervisor gets results through people. The supervisor is accountable for outcomes — production, cost, safety, quality, maintenance, training — but doesn't achieve those outcomes alone. The outcomes happen through the people the supervisor leads. The training visualizes a "Job Relations line" between the supervisor and the team. If the line is straight, strong, and correct, the chance of hitting the standard outcomes is high. If the line is long and wobbly, the outcomes are harder to achieve. The line always exists. The quality of the line is what varies.
The Improvement Kata, Oscar argued, has the same structure with a different orientation. The improvement in results happens through people. Target conditions don't get achieved by the coach. They get achieved by the learners — the doers practicing the experiments. There's a Job Relations line between the coach and the learners. If that line is straight, strong, and correct, the target conditions get achieved. If the line is long and wobbly, the experiments stall and the target conditions remain aspirational.
The four foundations of Job Relations apply equally well in both directions. The supervisor-to-worker relationship that supports stable performance and the coach-to-learner relationship that supports improvement both depend on the same proactive practices — let each person know how they're doing, give credit when due, tell people about changes that will affect them and explain why, make best use of each person's ability.
Oscar quoted Dave Hire, a director at the Boeing site, from a webinar a few months earlier: "If people on your team don't trust your intent, it'll be difficult to make progress." Hire had specifically identified Job Relations as a key skill in getting people to trust the leader's intent. Oscar's extension: a fair amount of step 1 of the Improvement Kata (understanding the direction or challenge) is about setting intent. If the team doesn't trust the intent, they won't buy into the direction, won't contribute meaningfully to understanding the current condition, won't engage in setting the target condition, and won't experiment toward it when the coach isn't there.
The experiments happen when the coach isn't there. The coaching cycle is a summary of what happened and a review of what's planned next. The actual work happens in between coaching cycles. For that work to actually happen as planned when the coach isn't present, the Job Relations line has to be straight, strong, and correct. Otherwise the experiment is something the doer agrees to in the meeting and doesn't actually do afterward.
The implication for organizations adopting Toyota Kata without TWI: the Improvement Kata can be taught, the coaching cycles can be practiced, the target conditions can be set — but if the underlying leadership relationships aren't supporting the doer's day-to-day work, the experiments won't run when the coach steps away. The skill of leading is what makes the skill of scientific thinking operationally viable.
Oscar closed with a framing of what a culture of coaching actually is, drawing on Stephen Covey's 1989 diagram of culture as the intersection of knowledge, skills, and attitude/desire.
Knowledge produces the what and why. Skills produce the how-to. Attitude and desire produce the want-to. Where all three intersect, you get habits, behaviors, and mindsets. Those habits, behaviors, and mindsets are the culture.
A Lean culture, Oscar argued, can be identified by the practice of the four fundamental skills — JI, JR, JM, and the Improvement Kata — as countermeasures to the symptoms that surface in daily work. But the skills alone don't produce culture. The skills plus a coach produce culture, because the coach is the person who builds knowledge, skills, and attitude/desire in the doer through the iterative coaching cycles each skill is built around.
The familiar continuous improvement diagram — the wedge under the wheel, holding the wheel in place against the slope of regression — fits inside this framing. The standard is held in place by workers who know how, can do, care, and do (JI plus JR). The wheel rolls uphill through goal-directed experiments (the Improvement Kata supported by JM). The coach is what keeps the system running rather than stalling.
Oscar closed with the same definition of Lean he'd opened with. Lean is developing people who can and will solve problems and make improvements daily. The waste elimination happens as a consequence. He'd never seen a sustained improvement or a sustained problem solved that hadn't eliminated waste along the way, but the waste reduction was never the point. The point was the people development. The point was the culture. The waste reduction was the predictable result.
The four skills Oscar walked through are conceptually independent of any software. The Job Instruction breakdown can be written on paper. The Job Relations four-step pattern can be applied in a conversation. The Job Methods six questions can be asked at the gemba. The Improvement Kata storyboard can be a whiteboard on the wall. The work doesn't require infrastructure to be possible.
What infrastructure does is preserve the practice when the practice has to operate at scale across many teams and over the kind of time horizons that real cultural change requires.
The coaching pattern at the heart of all four skills depends on continuity. The same coach and doer meeting repeatedly. Each cycle's learning informing the next. The track record of what was tried, what was learned, and what changed as a result accumulating into the kind of evidence that distinguishes a real coaching relationship from a series of disconnected conversations. When the coach is the doer's manager — which is what Oscar argued should be the case — the continuity is easier because the relationship already exists. When the coach is someone external to the doer's daily work, the continuity has to be supported deliberately. Either way, the artifacts of the coaching cycles need somewhere to live.
The pull-system relationship Oscar described — the Improvement Kata as the pull, the TWI skills as the countermeasures pulled in by surfaced obstacles — works best when the obstacles, the experiments, and the countermeasures can be tracked together. A team running Improvement Kata cycles on a whiteboard can do this for a single target condition. A whole organization running many Improvement Kata cycles across many teams toward many target conditions needs a way to keep the work visible across the organization. The improvements get spread when one team's countermeasure to an obstacle is visible to another team facing the same obstacle. The improvements get sustained when the post-experiment follow-up (the "Act" or "Adjust" in PDCA) actually happens because someone is tracking whether it happened.
The training schedule that Job Instruction depends on — the timetable of who needs to be trained on what, by when — is a piece of operational tracking that scales poorly in spreadsheets across a multi-site organization. The same is true for the Job Methods improvements that get implemented across many cells or many units, and need to be propagated so that one cell's improvement reaches the other cells doing the same work. Infrastructure that captures these as part of the broader improvement portfolio is what lets the practices spread rather than stay local.
The cultural shift Oscar was advocating for — from waste-focused improvement to people-development-focused improvement, with the coach-doer relationship at the center — is supported when the systems leaders use to manage improvement reflect that focus. A dashboard that only tracks waste eliminated reinforces the waste-elimination framing. A system that tracks the development of people, the practice of coaching cycles, the obstacles surfaced and addressed, and the target conditions achieved supports the people-development framing. The display itself becomes a reinforcement of the cultural direction.
None of this changes what Oscar was teaching. The skills are the skills. The symptoms are the symptoms. The PDCA backbone is what makes the skills coherent. The coaching pattern is what makes the practice cumulative. What infrastructure does is preserve the integrity of the practice when the practice is being applied across an organization rather than by a single skilled practitioner.
What's the working definition of Lean Oscar uses? Lean is developing people who can and will solve problems and make improvements daily. The definition reframes the focus from waste elimination (what most early Lean implementations emphasized) to people development. Waste reduction still happens — Oscar said he's never seen a sustained improvement that didn't eliminate waste along the way — but it happens as a consequence of the people development rather than as the goal.
What's the relationship between TWI and Toyota Kata? They're complementary skill patterns built on the same PDCA foundation. TWI provides three skills — Job Instruction, Job Relations, and Job Methods — that address specific symptoms in frontline work (not knowing how, not caring, physical difficulty). Toyota Kata provides the Improvement Kata, which addresses random or unfocused improvement. The four skills work together: the Improvement Kata creates the experimenting structure that surfaces obstacles, and the TWI skills are the countermeasures pulled in to address those obstacles when they appear.
Why does Oscar say the skills should be understood as countermeasures rather than as practices to adopt for their own sake? Because adopting a skill without a clear understanding of what symptom it addresses produces unsustainable adoption. The skill becomes overhead rather than a tool. When Oscar asks organizations why they want to start practicing a particular skill, the answers reveal whether the adoption has a real chance — if the organization can connect the skill to a specific symptom they're trying to address, the chances of sustained practice are much higher than if they're adopting the skill because it's in the air or because someone in leadership read a book about it.
Who should be the coach in a coaching relationship? Ideally, the doer's direct manager. The manager is the person closest to the doer's daily work, the person most invested in the doer's development, and the person most likely to be present when coaching opportunities arise. External coaches and internal CI departments have a role — Oscar said their role is to help managers establish the framework — but their job isn't to be the framework's coaches. When external coaches end up doing the coaching work because the managers aren't, the practice doesn't sustain after the external coaches leave.
What does "PDCA for human beings" mean in Job Relations? The four-step pattern in Job Relations for handling a people problem follows the same Plan-Do-Check-Act structure as the other three skills, applied to a human situation rather than a technical one. Get the facts (Plan). Weigh and decide (still Plan). Take action (Do). Check results (Check and Act). The pattern is the scientific method applied to relationships, which sounds counterintuitive but works because relationships have observable consequences and human behavior responds to interventions in patterns that can be predicted, tested, and adjusted.
Why do the six questions in Job Methods need to be asked in a specific order? Because the order is the point. Without the discipline of asking "why" first (and "where," "when," and "who" before "how"), supervisors jump straight to "how" — and end up optimizing the execution of steps that shouldn't exist at all, or that would be much better done somewhere else, at a different time, or by a different person. Oscar quoted Peter Drucker on the danger of doing something really well that shouldn't be done at all. The Job Methods question order is structured to surface elimination opportunities and relocation opportunities before optimization opportunities, which prevents organizations from getting very good at doing the wrong work.
What's the most important step of the Improvement Kata, in Oscar's view? Establishing the next target condition is the hardest part for most practitioners. The target condition isn't the goal. It's an interim state between current condition and the longer-term goal — a description of what the process needs to look like at a specific date in the future that, if achieved, will move the system closer to the goal. People naturally want to skip from current condition to solving problems. The discipline of stopping to describe a specific interim target condition is what makes the experiments goal-directed rather than random. Once the target condition is clear, the obstacles between current and target become visible, and the experiments aim at those obstacles rather than at whatever happens to be convenient.
What's the danger of "random or recreational" improvement? Two dangers. First, the improvements don't add up to anything coherent because they aren't aimed in a consistent direction. The organization is busy improving but not getting anywhere because the improvements don't compound. Second — and this came from Barbara in the Q&A — random improvement is often a strategy for avoiding the difficult problems. Practitioners pick the easy improvements because the easy improvements feel like progress, while the hard problems sit unaddressed. The Improvement Kata's discipline of setting a target condition first and then identifying which obstacles are in the path forces practitioners to address the hard obstacles rather than work around them.
Why is the Improvement Kata described as a pull system for the TWI skills? Because the obstacles that surface during the experimentation phase of the Improvement Kata are what create the demand for the TWI skills. The organization doesn't decide to "implement" Job Instruction. The organization runs an Improvement Kata cycle, surfaces an obstacle that's a knowledge gap, and pulls in Job Instruction as the countermeasure. The same is true for Job Relations (when the obstacle is engagement) and Job Methods (when the obstacle is physical difficulty). The pull is what creates sustained practice. Without it, the skills have to be pushed, and pushed practices rarely sustain.
What did Oscar mean about the "Job Relations line" in the Improvement Kata? That the same relational quality that determines whether a supervisor's team hits its standard outcomes also determines whether a coach's learner achieves the target condition. In Job Relations, the supervisor gets results through people, and the quality of the supervisor-to-worker relationship (the "Job Relations line") determines whether the results actually happen. In the Improvement Kata, the improvement happens through people. The coach gets target conditions achieved through the learner, and the quality of the coach-to-learner relationship determines whether the experiments actually run when the coach isn't there. The line always exists. The quality of the line is what varies, and the four foundations of Job Relations apply to building that quality regardless of which direction the line runs.
Can TWI and Toyota Kata be applied outside manufacturing? Yes. Oscar's business partner's main client is a county council in central New South Wales — a service organization whose customers are ratepayers. The skills apply directly. The Improvement Kata's structure — direction, current condition, target condition, experiments — works the same way regardless of whether the process produces bottles of wine, broadband installations, patient care, or municipal services. Goals will be framed differently, target conditions will look different, the specific obstacles will differ. The pattern is the same. Skip Steward at Baptist Memorial has done substantial work applying Job Instruction, Job Relations, and the Improvement Kata together in healthcare. Mark added that TWI was originally used in healthcare during World War II, and Virginia Mason Medical Center has written extensively about applying it in modern healthcare settings.
What's the role of external coaches if the manager should be the coach? The role of external coaches (or internal CI departments) is to help management establish the framework, not to be the framework's coaches. Oscar was emphatic on this. If external coaches end up doing the coaching work, the practice depends on the external coaches and stops when they leave. If external coaches help managers learn to coach — building the managers' skill in Job Instruction coaching, Job Relations coaching, Job Methods coaching, and Improvement Kata coaching — the practice transfers to the organization and continues after the external resources withdraw. Oscar's framing: who's responsible for the doer's development? The doer's manager. Who's responsible for the manager's development? The manager's manager. The chain runs through the organization rather than around it.
What's the connection to Stephen Covey's culture model? Oscar used Covey's 1989 framing of culture as the intersection of knowledge (what and why), skills (how-to), and attitude/desire (want-to). Where all three intersect, habits, behaviors, and mindsets emerge. Those habits, behaviors, and mindsets are the culture. The coach's role in TWI and Toyota Kata is to build knowledge, skills, and attitude/desire in the doer through the iterative coaching cycles each skill is built around. The four skills plus a coach is what produces the culture. Skills without a coach is just training. A coach without skills is just supervision. The combination is what shifts culture.
What did Mark and Oscar reference about Dave Hire and Boeing? Dave Hire is a director at a Boeing site who had presented in an earlier Lean Frontiers webinar. The line Oscar quoted: "If people on your team don't trust your intent, it'll be difficult to make progress." Hire specifically identified Job Relations as a key skill in building the trust that makes progress possible. Oscar used the quote to bridge from Job Relations into the Improvement Kata — step 1 of the Improvement Kata is partly about setting intent, and if the team doesn't trust the intent, the rest of the kata stalls.
What resources does Oscar recommend for learning more? The session referenced David Meier's "Toyota Talent" (specifically the line on key points being the most important aspect of training), Mike Rother's "Toyota Kata" and the associated practitioner materials, the Stephen Covey culture model, and Skip Steward's work at Baptist Memorial as an example of TWI and Kata combined in healthcare. The TWI Institute website (twi-institute.org) is the central resource for the TWI side of the work. Lean Frontiers (leanfrontiers.com) runs the annual TWI Summit and Toyota Kata Summit, where Oscar regularly presents and runs simulations that illustrate the skills in action. Mark also mentioned the Virginia Mason Medical Center book on standard work in healthcare as a reference for the healthcare application.
.svg)
Copyright © 2026
Privacy Policy