Kata
Kata is a routine practised until it becomes a habit. In lean management, it consists of two linked routines from Toyota. The improvement kata follows four steps to reach a target condition. Teams set this target condition one to four weeks out. A learner experiments toward it daily, addressing one obstacle at a time. The coaching kata consists of five standardized questions. A coach asks these five questions daily in the exact same order. Both routines are tracked on a storyboard with five fields.
- Challenge card for valve line
- This storyboard hangs on the V-220 valve line. The focus process is the LT-3 leak tester. The line runs at a 40-second takt to produce 90 valves an hour. The challenge requires every hour at plan, every shift, by year end. It is too far away to solve directly. It gives direction to every target condition.
- Measured current condition card
- The current condition is measured, not estimated. On day 0, the tester averaged 4 timeouts per shift. The line achieved only 5 of 8 hours at plan. The learner remeasures this condition after every coaching cycle. The numbers update as the record grows. By cycle 8, the condition reaches 0 timeouts and 8 of 8 hours.
- Target condition for process
- A target condition describes how the process should operate on a specific date. The window is one to four weeks out. This card sets a two-week target for Friday 26 September. The target requires at most 1 timeout per shift and 8 hours at plan. It is not a solution. The short window forces fast experiments instead of a plan.
- Obstacle parking lot list
- Obstacles stand between the current condition and the target condition. Four obstacles sit in this lot. The learner addresses one obstacle at a time. The rest wait without work. An obstacle is struck through only when data shows it no longer causes timeouts. In this study, evidence closed all four listed obstacles across eight coaching cycles.
- Eight coaching cycle results
- One coaching cycle covers one shift, one step, and one measurement. Timeouts per shift were 4, 2, 3, 1, 1, 2, 0, 0. The count rose twice on the way down. Each step tested a separate cause. The learner reached the target condition on cycle 7. Cycle 8 held the result two days before the target date.
- Cycle experiment learning lines
- Every cycle records step, expectation, what happened, and what was learned. The learner writes the expectation before taking the step. The learner then compares the outcome with the prediction. A wrong prediction is a finding, not a failure. This comparison trains scientific thinking. The routine turns each experiment into a Plan-Do-Check-Act cycle on the board.
- Five coaching kata questions
- The coaching kata routine lists five questions. The coach asks them in the same order every shift. The questions cover target condition, actual condition, current obstacle, next step with expectation, and when to follow up. S.L. answers them daily for R.K. at the storyboard. This strict routine trains the coach as much as the learner.
Key facts
- Japanese term
- 型 — a form practised until it is routine
- Two routines
- Improvement kata (learner) and coaching kata (coach)
- Four steps
- Direction, current condition, next target condition, experiments
- Target condition window
- 1 to 4 weeks; this storyboard uses 2
- Coaching cycles on the board
- 8, one per shift
- Leak-tester timeouts per shift
- 4 to 0; target was at most 1
By Matthew Savas — Founder of Kaizumi. Reviewed 13 September 2026.
Kata is a routine practised until it becomes habit. In lean management it names two linked routines from Toyota. The improvement kata moves a process from its current condition to a target condition set one to four weeks out, one obstacle at a time. The coaching kata is the five questions a coach asks the learner every day in the same order.
Together, these routines teach scientific thinking on the shop floor. Teams use them to bridge the gap between current performance and strategic operational requirements.
What is kata?
The Japanese word kata (型) refers to a detailed choreographed pattern of martial arts movements. Practitioners repeat these patterns until the execution becomes an unconscious, automatic habit.
In production management, kata refers to deliberate behavioral routines. These routines train people to solve operational problems systematically. Continuous improvement requires systematic habits rather than occasional workshops. Kata embeds daily scientific experimentation into frontline work.
The framework consists of two core components: the improvement kata and the coaching kata. A learner practices the improvement pattern to address operational obstacles. A manager practices the coaching pattern to guide the learner. Daily practice creates a shared approach to problem solving across the plant.
Origins in Toyota management practice
Mike Rother introduced the terminology in his 2009 book Toyota Kata. Rother studied Toyota factories to understand how the company sustained performance gains over decades. He found that specific tools like kanban did not explain Toyota's continuous gains.
Instead, Toyota managers relied on daily habits of coaching and iterative experimentation. Rother codified these habits into the improvement kata and the coaching kata. The improvement kata defines the learning pattern for frontline operators and engineers. The coaching kata defines the supervisory routine for managers and executives.
Both routines shift operational culture away from guessing and predetermined implementation plans. They train employees to navigate uncertain territory through structured, empirical steps.
The four steps of the improvement kata
The improvement kata follows four sequential steps:
- Understand the direction or challenge. The organization defines an overarching operational need. This challenge comes from annual business objectives or customer requirements. It provides the long-term direction.
- Grasp the current condition. The learner goes to the gemba to observe the process directly. The learner documents factual process metrics, cycle times, defects, and work patterns.
- Establish the next target condition. The team defines a detailed operating pattern to achieve in one to four weeks. This step sets an intermediate milestone between the current condition and the challenge.
- Experiment toward the target condition. The learner tests countermeasures against specific obstacles using iterative cycles. The learner records expectations before taking action and evaluates physical outcomes after each test.
Learners repeat step four daily until the process reaches the target condition. Once achieved, the team grasps the new current condition and establishes another target condition.
Target conditions compared to goals and solutions
A target condition is not a numeric goal alone. It is also not an action plan or an implementation checklist. A goal states a desired outcome metric, such as defect rates or line throughput. A solution predetermines a mechanical or procedural change before testing.
A target condition describes how the process should operate on a specific future date. It defines the operating pattern, process characteristics, cycle time, and measurable outcomes together. For example, a goal might state a requirement to produce 90 units per hour. A target condition specifies takt time, operator work sequence, station work-in-process, and scrap limits.
Setting a target condition defines the operational boundary for daily experimentation. When a process fails to run in that described pattern, an obstacle reveals itself. The learner then addresses that specific obstacle through direct physical testing.
The coaching kata and the five questions
The coaching kata develops the learner's scientific thinking. A coach, typically a supervisor or group leader, meets with the learner every working day. The coaching session takes place at the process storyboard on the shop floor.
The coach asks five standardized questions in the exact same order every day:
- What is the target condition?
- What is the actual condition now?
- What obstacles are in the way, and which one are you addressing now?
- What is your next step, and what do you expect?
- When can we go and see what we have learned?
The coach evaluates the learner's logic rather than providing solutions. If the learner jumps to conclusions without direct observation, the coach redirects the learner. This daily rhythm establishes leader standard work centered on capability development. Print the coaching kata card to follow the five questions during daily cycles.
The five fields of the storyboard
The learner maintains an improvement kata storyboard near the physical workstation. The storyboard serves as a dynamic workspace rather than a reporting display. It contains five distinct fields:
- Challenge: The high-level strategic operational goal aligned with executive planning.
- Current condition: Process metrics, actual cycle times, work sequence, and scrap rates recorded at the gemba.
- Target condition: The desired operational pattern, metrics, and date, established one to four weeks ahead.
- Obstacle parking lot: A documented list of identified barriers preventing the target condition. The learner selects only one obstacle to address at a time.
- PDCA record: An experimental log containing four columns: step, expectation, what happened, and what was learned.
The learner updates the storyboard prior to each daily coaching session. The coach uses the board to inspect the learner's evidence and line of reasoning. Use the Toyota Kata storyboard to manage the five fields directly inside your browser.
Kata, kaizen, PDCA, and A3 problem solving
Kata connects closely with other operational improvement concepts while fulfilling a distinct role. Kaizen represents the philosophy and outcome of continuous operational improvement. Kata is the specific behavioral practice routine that produces kaizen.
Each experimental step on the storyboard corresponds to one PDCA cycle. The Plan phase defines the step and explicit expectation. The Do phase executes the physical test. The Check phase records what happened. The Act phase documents what was learned and standardizes successful adjustments. For further technical details on running these cycles, see the PDCA guide.
A3 problem solving addresses broader, systemic operational issues across departments. An A3 report synthesizes background analysis, root causes, countermeasures, and schedules onto a single A3 sheet. In contrast, kata serves as a daily, rapid learning habit on the production line. A team might use an A3 project to define a major challenge, then use kata routines to work toward it daily.
Example: two weeks at a leak tester
The V-220 valve assembly line operates to a 40-second takt time. Its target output is 90 units per hour. The line displays output on an hourly production board. Over past shifts, "leak-tester timeout" appeared repeatedly as the primary cause for missed output.
The focus process is the LT-3 leak tester at the end of the line. The line challenge requires every hour at plan, every shift, by year end. S.L., the team leader, acts as the learner. R.K., the group leader, acts as the coach.
On day zero, Monday 15 September, S.L. grasps the current condition. The line experiences 4 tester timeouts per shift, achieving plan in only 5 of 8 hours. S.L. and R.K. establish the target condition for Friday 26 September: at most 1 timeout per shift and 8 of 8 hours at plan.
S.L. lists four obstacles in the obstacle parking lot:
- Nobody knows when the timeouts happen.
- The tester cools down over breaks.
- The valve is not fully seated on the test port.
- Air pressure dips when the press cycles.
S.L. runs eight daily coaching cycles, recording each on the PDCA record.
In cycle 1 on Monday 15 September, S.L. logs the clock time of every timeout for one shift. S.L. expects timeouts spread across the shift. Instead, 4 timeouts occur, with 3 in the first cycle after a break. S.L. learns timeouts cluster after breaks, not at random. The condition remains 4 timeouts, 5 of 8 hours.
In cycle 2 on Tuesday 16 September, S.L. leaves the tester powered through the two breaks. S.L. expects timeouts to drop to 1 or 2. The tester records 2 timeouts, both mid-hour, neither after a break. S.L. learns cooling over breaks caused 2 of the 4 timeouts. The condition improves to 2 timeouts, 6 of 8 hours.
In cycle 3 on Wednesday 17 September, S.L. watches 20 tests at the port and notes seating. S.L. expects timeouts follow a valve seated at an angle. S.L. observes 3 timeouts, every one on a valve seated at an angle. S.L. learns seating is the second cause; the port lacks a stop. The condition reads 3 timeouts, 6 of 8 hours.
In cycle 4 on Thursday 18 September, S.L. fits a temporary cardboard locating stop on the test port. S.L. expects zero timeouts from seating. The shift logs 1 timeout; the stop crushed by the afternoon. S.L. learns the stop works, but cardboard cannot survive a shift. The condition reads 1 timeout, 7 of 8 hours.
In cycle 5 on Friday 19 September, S.L. replaces the cardboard stop with a machined aluminum stop. S.L. expects zero seating timeouts for the whole shift. The shift records 1 timeout at 14:10, valve seated square. S.L. learns seating issues are closed; the final timeout has another cause. The condition remains 1 timeout, 7 of 8 hours.
In cycle 6 on Monday 22 September, S.L. fits a pressure gauge at the tester and reads it at timeouts. S.L. expects air pressure dips below 5 bar when the press cycles. S.L. logs 2 timeouts; the gauge reads 4.6 bar at both occurrences. S.L. learns supply air pressure is the last cause, as the press shares the drop. The condition reads 2 timeouts, 7 of 8 hours.
In cycle 7 on Tuesday 23 September, S.L. fits a 10-liter air receiver on the tester drop. S.L. expects pressure holds above 5 bar through a press cycle. The tester logs 0 timeouts; the gauge never drops below 5.4 bar. S.L. learns the receiver holds pressure and plans a verification run. The condition reaches 0 timeouts, 8 of 8 hours.
In cycle 8 on Wednesday 24 September, S.L. runs a second full shift with all three changes in place. S.L. expects at most 1 timeout and 8 hours at plan. The line logs 0 timeouts and 8 of 8 hours at 90. The target condition is achieved two days ahead of schedule.
Across the eight cycles, timeouts per shift measured: 4, 2, 3, 1, 1, 2, 0, 0. The count went up twice on the way down. The three permanent changes enter the standard work and the daily maintenance standards: tester powered through breaks, aluminum locating stop, and 10-liter air receiver. S.L. and R.K. then establish the next target condition from the new current condition.
Frequently asked questions
- How does a target condition differ from an operational goal?
- An operational goal defines a desired numeric result alone, such as reaching a specific throughput volume or defect rate. A target condition describes the detailed operating pattern required on a future date to achieve those results, including takt time, operator work sequences, and work-in-process limits. Describing how the process should run establishes a clear baseline that exposes operational obstacles during daily work.
- What is the primary role of the coach during the coaching kata?
- The coach evaluates the learner's scientific reasoning and evidence rather than offering solutions or action items. By asking five standardized questions at the storyboard, the coach ensures the learner bases decisions on direct gemba observation and empirical testing. If the learner jumps to conclusions without supporting data, the coach redirects them back to disciplined observation.
- Why does the improvement kata focus on only one obstacle at a time?
- Addressing a single obstacle isolates cause and effect during daily PDCA experimentation. If a learner tests multiple process changes simultaneously, the team cannot verify which specific adjustment produced the observed outcome. Remaining barriers are kept on the storyboard's obstacle parking lot until the active experiment concludes.
- What information does a learner document in the storyboard's PDCA record?
- The PDCA record tracks daily experimentation across four columns: step, expectation, what happened, and what was learned. Before taking action, the learner documents the planned physical test and its expected result. After executing the test at the process, the learner records the actual outcome and the operational insight gained.
- How does kata relate to kaizen and A3 problem solving?
- Kaizen is the overarching philosophy and result of continuous improvement, while kata provides the deliberate daily practice routine that generates that improvement. A3 problem solving typically tackles broader, cross-departmental challenges on a single summary sheet. Organizations often use an A3 project to define a major operational challenge, then deploy kata routines for daily line-level experimentation toward it.
Sources and notes
- Mike Rother, Toyota Kata (McGraw-Hill, 2009) — The improvement kata's four steps and the coaching kata's five questions
- Mike Rother, The Toyota Kata Practice Guide (McGraw-Hill, 2018) — The learner's storyboard: challenge, current condition, target condition, obstacle parking lot, PDCA record