"Kaizen means improvement. Moreover, it means continuing improvement in personal life, home life, social life, and working life." — Masaaki Imai
Activation Conditions and Trigger Words
- Direct invocation: "Use kaizen analysis..." "continuous improvement suggestions"
- Semantic triggers: the user mentions low efficiency, unstable quality, lengthy processes, or wanting gradual optimization
- Combined invocation: "First use kaizen to find optimization points, then use systems thinking to look at the overall structure"
Method Framework Overview
The core of kaizen is to accumulate major transformation through small-step, quick-iteration gradual optimization. Through genchi genbutsu (going to the actual place) to see real problems, standardization as the foundation, and eliminating the seven wastes, it enables anyone to continuously contribute improvements.
Core Principles (3-7, each must be accompanied by ≥2 pieces of cross-domain evidence)
Principle 1: Small-step improvements accumulate into major transformation
One-sentence definition: Improve 1% every day, and after a year you will be 37 times stronger.
Cross-domain evidence：
- Automotive manufacturing: Toyota receives millions of employee improvement suggestions each year, dozens per person per year, continuously accumulating into TPS (Source: Liker, The Toyota Way, 2004)
- Animation production: At Pixar's daily stand-up meetings (Dailies), everyone can propose improvement suggestions, and each film is iterated hundreds of times (Source: Catmull, Creativity Inc., 2014)
How to apply: Do not create a perfect improvement plan; instead, start today by doing one small thing that can improve things. Record what improvement you did each day.
Limitations: Small-step improvement is not enough for scenarios that require fundamental breakthroughs. Structural problems require systemic change rather than incremental improvement.
Principle 2: Genchi Genbutsu — go to the actual place to see real problems
One-sentence definition: You cannot see the real problems by looking at data at your desk. Go to the actual place where the problem occurs.
Cross-domain evidence：
- Automotive manufacturing: Toyota management is required to "stand on the production line and observe for 4 hours" every week (Source: Ohno, Toyota Production System, 1988)
- Technology management: Amazon's Bezos insists on regularly working on the warehouse front line to discover delivery pain points (Source: Stone, The Everything Store, 2013)
How to apply: Go directly to the actual place where the problem occurs — the production line, customer service call recordings, screen recordings of users using the product — rather than looking at PPT reports.
Limitations: On-site observation is sampling-based and needs to be combined with large-scale data for judgment.
Principle 3: Standardization is the cornerstone of kaizen
One-sentence definition: Without a standard, there is no kaizen — if the method is different every time, you cannot tell whether an "improvement" is a real improvement or random fluctuation.
Cross-domain evidence：
- Food service: McDonald's standardized operating procedures enable every store worldwide to make a consistent hamburger, and then iterate and optimize on that standard (Source: Love, McDonald's: Behind the Arches, 1986)
- Software development: CI/CD standardized processes enable teams to measure the speed and quality of each deployment, and continuously improve on that basis (Source: Forsgren, Humble, Kim, Accelerate, 2018)
How to apply: First write down the current process (even if it's imperfect); this is the starting point for improvement. Then ask: which step is the slowest? Which step is most prone to errors?
Limitations: Over-standardization may stifle creativity and flexibility. Standardization applies to repetitive processes, not to creative work.
Principle 4: Eliminating waste (Muda) is the core goal
One-sentence definition: Any activity that does not add customer value is waste.
Cross-domain evidence：
- Manufacturing: Toyota identified 7 types of waste: overproduction, waiting, transportation, over-processing, inventory, motion, defects (Source: Ohno, Toyota Production System, 1988)
- Software development: Lean Software Development translated the 7 types of waste into the software context: unfinished features, waiting for approval, manual work, etc. (Source: Poppendieck, Lean Software Development, 2003)
How to apply: List everything you do in a day, and ask one by one, "Would the user or customer pay for this?" If not—it's waste; find a way to eliminate or reduce it.
Limitations: Some activities that "do not directly add customer value" are necessary (such as compliance checks and security testing). Not all non-value-adding activities are eliminable waste.
Operating Protocol (Agentic Protocol)
Step 1: Problem classification
Applicable signals：
- Efficiency and quality problems in repetitive processes
- The team continuously feels inefficient but doesn't know where to start improving
- Wanting to build a culture of improvement but not knowing where to begin
Not applicable signals：
- Need for fundamental business model change (improvement is incremental)
- "Efficiency" problems in creative workflows (creativity is hard to standardize)
- Urgent crisis requiring a quick solution
Step 2: Improvement-style analysis
Research dimensions(each dimension must be traced back to a core principle):
-
Waste identification: In the current process, which things are customers not paying for?
- Source principle: Principle 4 (Eliminate waste)
- How to do it: Scan one by one according to the 7 types of waste, and write down specific examples
-
Gemba observation: Directly observe the place where the problem occurs rather than looking at data reports
- Source principle: Principle 2 (Gemba-ism)
- How to do it: If there is a physical site, go there; if it is a digital product, watch user recordings or customer service call recordings
-
Review of current standardization: Does the current process have a standard?
- Source principle: Principle 3 (Standardization)
- How to do it: Write down the current most "standard" practice. If there is no standard—that is the first improvement
-
Prioritizing improvement opportunities: Find the 3 places with the greatest waste, and start with the smallest and easiest one to change
- Source principle: Principle 1 (Small-step improvement)
- How to do it: Prioritize using an "impact × feasibility" matrix, and do the high-impact + high-feasibility ones first
Step 3: Improvement-style output
- Diagnostic results for the 7 types of waste
- Missing points in the current process standard
- Top 3 improvement opportunities (ranked by impact × feasibility)
- The first improvement action (the smallest change that can be done today)
Applicable/not applicable judgment
| Problem type | Applicability | Description |
|---|---|---|
| Efficiency optimization of repetitive processes | High | Classic scenario |
| Reduction of quality defect rate | High | Kaizen excels at defect elimination |
| Daily optimization of team operations | High | Suitable for building improvement habits |
| Business model innovation | Low | Kaizen is an improvement model, not an invention model |
| Extremely urgent crisis handling | Low | Kaizen requires time and a system, and is not suitable for purely urgent scenarios |
Classic case library
Success cases
-
Toyota Production System(Automotive manufacturing/1970s to present)
- Problem: Postwar Japan lacked resources and could not afford waste
- Method application: Establish an all-employee improvement system, where every worker can pull the cord to stop the production line to solve quality problems
- Result: Became the world's most efficient and most reliable automotive manufacturing system
- Source: Ohno, Toyota Production System (1988); Liker, The Toyota Way (2004)
-
Pixar daily improvement stand-up meetings(Animation/2000s)
- Problem: Animation production is lengthy, and errors are discovered late
- Method application: Daily Dailies stand-up meetings—anyone can offer improvement suggestions for any frame
- Result: Each film undergoes hundreds of small improvements, with quality far above the industry average
- Source: Catmull, Creativity Inc. (2014)
-
Amazon delivery optimization(E-commerce/2000s to present)
- Problem: High delivery costs and many errors
- Method application: Under Bezos's leadership, continuously optimize every link in the delivery process—warehouse layout, picking routes, packaging materials
- Result: The Prime membership experience became an industry benchmark
- Source: Stone, The Everything Store (2013)
Failure cases
- False implementation of Kaizen(Management/General)
- Problem: The company announces "We are going to do improvement!" but it remains only a slogan
- Misuse method: Hang banners, shout slogans, ask employees for suggestions but do not implement them
- Lesson: Kaizen requires a system—a closed loop of suggestion → review → execution → feedback. Collecting without executing is worse than doing nothing
- Source: Imai, Gemba Kaizen (1997); numerous management practice reports
Misuse detector (≥3 misuse patterns)
| Misuse signal | Detection logic | Warning message | Recommended action |
|---|---|---|---|
| Method and problem do not match | The user needs breakthrough innovation but wants to use Kaizen | "Kaizen is incremental improvement, not breakthrough innovation. What you need is first principles or TRIZ, not Kaizen." | Recommend alternative methods |
| Skipping key steps | The user wants to directly "optimize" but has not yet standardized the current process | "Without a standard, there is no improvement. First write down the current process; this is the baseline for improvement." | Guide back to standardization |
| Slogan-style implementation | The user says "all-employee improvement" but has no suggestion review and execution mechanism | "Just saying 'everyone give suggestions' without closed-loop management is worse than doing nothing. You need to establish a complete closed loop of suggestion → review → execution → feedback." | Guide the establishment of a closed loop |
Honest boundaries (≥3 specific limitations)
- Not suitable for scenarios requiring breakthrough innovation: Kaizen makes good better, but it cannot turn "wrong" into "right". Kaizen in the wrong direction only accelerates the march toward wrongness
- Speed limit: The incremental nature of Kaizen means it takes time. If your problem must be solved within 24 hours, Kaizen is too slow
- Cultural dependency: The success of Kaizen depends on a culture where "everyone can offer opinions". In a rigidly hierarchical or fear-based culture, Kaizen cannot truly operate
- Standardization may become inertia: Excessive standardization may lead teams to "follow the rules" without questioning whether the standards themselves are reasonable
Research sources (proportion of primary sources must be >50%)
| # | Source | Type | Primary/Secondary | Weight |
|---|---|---|---|---|
| 1 | Ohno, Toyota Production System (1988) | Book | Primary | High |
| 2 | Imai, Kaizen (1986) | Book | Primary | High |
| 3 | Liker, The Toyota Way (2004) | Book | Primary | High |
| 4 | Imai, Gemba Kaizen (1997) | Book | Primary | Medium |
| 5 | Catmull, Creativity Inc. (2014) | Book | Primary | Medium |
| 6 | Poppendieck, Lean Software Dev (2003) | Book | Primary | Medium |
| 7 | Stone, The Everything Store (2013) | Biography | Secondary | Medium |
This Skill was generated by Forge Skill — Forging Thinking Tools generated