The Improvement Kata is a four-step scientific routine, popularized by Mike Rother in his 2009 book Toyota Kata: understand the direction, grasp the current condition, set the next target condition, then run experiments toward it. Practiced daily, it turns problem-solving into a habit rather than a one-off event.
Most plants improve in bursts. A kaizen event fires up, gains hold for a month, then the floor drifts back. The Improvement Kata attacks that drift by changing how people think every day, not just during events. It is a deliberate practice routine, the word kata comes from a rehearsed form in martial arts, that trains a scientific way of working toward a goal you cannot yet see the whole path to. It sits at the behavioral core of lean manufacturing: the tools matter less than the daily habit of setting a target and experimenting your way to it.
What Is the Improvement Kata?
The Improvement Kata is a structured, repeatable routine for moving a process from where it is toward a challenging goal through small, fast experiments. Instead of planning the whole solution up front, you pick the next short-term target condition, then run experiments one after another, letting each obstacle you hit tell you what to work on next. It is discovery, not prediction.
Mike Rother, an American researcher, named and codified the pattern after studying how Toyota manages improvement. His finding was that Toyota's edge was less about specific tools and more about a shared, teachable routine, and a matching coaching routine, that builds scientific thinking into daily management. The point is the practice, not any single result: you run the kata to get better at getting better.
Why Is It Called a "Kata"?
A kata is a rehearsed form you practice until it becomes second nature, the way a musician runs scales or a martial artist drills a sequence. The name is deliberate. You are not memorizing an answer; you are grooving a way of thinking, so that under pressure your default is to observe, set a target, and experiment rather than to jump to a favorite solution. Repetition is the point. The first dozen times through the routine feel clunky; that is what practice is for.
This is why the Improvement Kata is a daily behavior, not a workshop. A one-week event teaches a tool. A kata practiced a few minutes every day, with a coach, teaches a habit. The distinction is the whole idea: Rother's argument is that lasting capability comes from converting lean tools into repeated behavior, so improvement continues after the consultants leave.
What Are the Four Steps of the Improvement Kata?
The routine is four steps, run in order and then repeated. The first is a longer-horizon setup; the last three cycle rapidly.
- Understand the direction or challenge. Start from a longer-term challenge tied to the plant's needs, "cut changeover to under 10 minutes," "one-piece flow on line 3." This is the true north the whole effort points at, set by leadership, not invented at the machine. It gives every experiment a reason.
- Grasp the current condition. Go and see. Measure how the process actually runs right now, with real data and direct observation, not opinions or the way the work instruction says it should go. You cannot set an honest target until you know the real starting point, and this step almost always corrects a wrong assumption.
- Establish the next target condition. Define a specific, measurable condition you want the process to reach by a near date, typically one to four weeks out, not the final challenge. A good target condition describes how the process should operate and by when, and it should sit just beyond what you currently know how to achieve, so reaching it requires learning.
- Experiment toward the target condition. Run rapid PDCA cycles, plan, do, check, act, one experiment at a time. Each experiment tests one idea against the target, and each obstacle you hit becomes the focus of the next experiment. You iterate until the target condition is met, then return to step 3 and set the next one toward the challenge.
What Is the Coaching Kata?
The Improvement Kata is paired with a Coaching Kata because a beginner cannot reliably self-correct a new thinking pattern. A coach meets the learner daily at the process and asks a fixed set of five questions: (1) What is the target condition? (2) What is the actual condition now? (3) What obstacles are in the way, and which one are you tackling now? (4) What is your next step or experiment, and what do you expect? (5) When can we go and see what we learned? The learner answers; the coach does not solve the problem for them.
The five questions are themselves a kata, drilled until the coach asks them naturally. Their purpose is not to extract a status update but to keep the learner practicing the scientific pattern: state where you are going, where you are, what is blocking you, and what you will test next. Good coaching is why the routine sticks, and it is what makes the habit spread from one person to a team.
How Is the Improvement Kata Different From a Kaizen Event?
Both improve processes, but they run on different clocks and build different things. A kaizen event is a concentrated burst, a cross-functional team fixing a defined problem in a few focused days. It is excellent for a known problem with a knowable solution, and it delivers a visible jump. The Improvement Kata is a slow daily drip aimed at a challenge whose solution is not yet known, and its main product is capability, people who now think scientifically, as much as the process gain itself.
They are not rivals. Many plants use kaizen events for big set-piece breakthroughs and the Improvement Kata for the daily grind toward challenges no single event can crack. The kata also strengthens what events leave behind: the experimental mindset that keeps gains from eroding. Where a countermeasure needs formal follow-through, it feeds your corrective and preventive action process; where it uncovers a stubborn defect, the kata's experiments pair naturally with disciplined root cause analysis.
How Do You Start Practicing the Improvement Kata?
Start small and real. Pick one process, one learner, and one coach, and hold a short daily cycle, often fifteen minutes at the process, using a simple storyboard that shows the challenge, current condition, target condition, and the last and next experiment. Resist the urge to roll it out plant-wide on day one; the routine has to be learned by doing, one pair at a time, before it can spread. Tie the target conditions to something leadership actually cares about so the practice earns its place, and connect the daily boards to your improvement huddles so obstacles surfaced in the huddle become kata targets. Over months, the goal is not a stack of solved problems but a floor full of people who reach for observation and experiment by reflex.
The Improvement Kata by the numbers
The routine is well documented and freely described. The Lean Enterprise Institute defines Kata as two linked routines, the Improvement Kata and the Coaching Kata, that build scientific-thinking habits into daily work, with the Improvement Kata running in four stages: understand the direction, grasp the current condition, establish the next target condition, and PDCA toward it (Lean Enterprise Institute, Kata). The pattern was named and popularized by researcher Mike Rother in his 2009 book Toyota Kata, based on study of how Toyota manages improvement and coaching (Lean Enterprise Institute, on Rother's Toyota Kata). The Coaching Kata's five questions and the emphasis on daily practice at the process are the mechanism by which the thinking, not just a result, is transferred.
The routine lives or dies on honest current conditions. Step 2, grasp the current condition, demands real numbers about how the process runs right now, and step 4 demands you check each experiment's actual result against what you predicted. When those numbers come from memory or end-of-shift paper, the learner argues about the baseline instead of running experiments. Plants that capture cycle times, stops, and output automatically at the process give every kata pair a trustworthy current condition and a same-day read on each experiment, which is the everyday value of live floor data over your existing systems no rip-and-replace. For how one plant built a faster improvement loop, see our CLS case study.