Prompts / Operations & Project Management

ChatGPT prompt for kaizens

This prompt produces a focused kaizen plan that a team can run without guessing who does what. Use it when a recurring process has waste, delay, errors, or avoidable effort.

PromptOpen ChatGPTOpen Claude
Help me create a practical kaizen for this operational issue.

Process and problem: [PROCESS AND PROBLEM]
Current evidence: [CURRENT-STATE EVIDENCE]
Target condition: [TARGET CONDITION]
Team and constraints: [TEAM AND CONSTRAINTS]
Time available: [TIME AVAILABLE]

First, ask up to 3 clarifying questions only if a required input is missing. Do not assume root causes, baseline data, authority, or available tools that I have not provided.

Produce a one-page kaizen plan with these sections:
1. Problem statement: define the gap in observable terms, including where it occurs, how often, and its effect.
2. Current state: map the relevant steps from trigger to handoff; identify waiting, rework, approvals, queues, motion, or information gaps using only my evidence.
3. Likely causes: separate confirmed causes from hypotheses to test. Use a short 5 Whys chain only where the evidence supports it.
4. Countermeasures: propose the smallest reversible changes first. For each, name an owner, exact action, due date, required support, expected effect, and leading measure.
5. Experiment plan: state the baseline, success threshold, test duration, and how the team will capture results.
6. Follow-through: give a daily check-in cadence, a review date, and a standard-work update step if the change succeeds.

Use a table for the action plan. Keep the scope narrow enough to complete within [TIME AVAILABLE]; defer larger system changes to a clearly labeled next kaizen. Before answering, check that every action has one accountable owner and a measurable outcome, and flag claims or causes not supported by the evidence.

Fill in

PlaceholderWhat to enterExample
[PROCESS AND PROBLEM]Describe the workflow, where it breaks down, and the operational consequence.Customer refund requests sit in an inbox for 3 to 6 days before review; customers send repeat emails.
[CURRENT-STATE EVIDENCE]Paste observed timings, counts, examples, process steps, or team notes.42 requests last month; 31 lacked an order number; support copies details into a sheet, then finance approves twice daily.
[TARGET CONDITION]State the desired measurable condition and any quality or safety requirement.A complete refund request receives a decision within one business day, with no increase in incorrect refunds.
[TEAM AND CONSTRAINTS]List the people involved, their roles, decision limits, systems, and constraints.Two support specialists, one finance approver, Zendesk and Google Sheets; no new software this quarter.
[TIME AVAILABLE]Enter the practical time window for the improvement work and trial.Five working days to test changes, then a 30-minute review.

How to use

  1. Paste an observed problem and the current-state facts, not a theory of what caused it.
  2. Check that the baseline and target use the same unit, such as hours, defects, or requests.
  3. Assign named owners before sharing the plan; a role without a person is not ownership.
  4. After the test, send: “Here are the results against the baseline. Update the plan and recommend standard work or the next experiment.”

Variations

Workshop agenda

Use this when you need to facilitate a short kaizen session.

Variation
Create a 90-minute kaizen workshop agenda for [PROCESS], attended by [PARTICIPANTS], to address [PROBLEM]. Use the available evidence: [EVIDENCE]. Include timed activities for defining the problem, mapping the current state, identifying waste, selecting one testable countermeasure, assigning owners, and agreeing on measures. Produce a facilitator script, materials list, and a decision log template. Do not claim causes are proven unless the evidence demonstrates them. Ask up to 3 questions only if a required input is missing.

A3 summary

Use this when leadership needs an A3-style improvement brief.

Variation
Write an A3-style improvement summary for [PROCESS PROBLEM] using [EVIDENCE] and [PROPOSED CHANGE]. Include background, current condition, target condition, root-cause analysis with confirmed facts separated from hypotheses, countermeasures, implementation owners, timing, and follow-up measures. Keep it to one page of plain language. Flag missing baseline data or unsupported causal claims instead of filling gaps. Ask up to 3 clarifying questions only when an essential input is absent.

Daily huddle

Use this when a kaizen is already running and the team needs a huddle format.

Variation
Build a 10-minute daily kaizen huddle template for [IMPROVEMENT GOAL] during [TEST PERIOD]. Use these actions and measures: [ACTION LIST AND DATA]. Include yesterday’s result versus baseline, barriers, owner commitments due today, escalation criteria, and one field for learning that changes the experiment. Produce a reusable agenda and a filled example using only supplied facts. Do not turn a single day’s result into a conclusion. Ask up to 3 clarifying questions if needed.

Tips

  • Start with one failure mode, such as incomplete intake data or a delayed approval, rather than trying to repair an entire department.
  • Measure the condition at the point of work; an average monthly dashboard can hide the queue or handoff causing the delay.
  • Treat a countermeasure as an experiment until it improves the chosen measure without creating a quality, safety, or workload tradeoff.
  • Update standard work only after the trial is stable enough to teach and audit.

FAQ

What is the difference between a kaizen and a project plan?

A kaizen tests a small, specific improvement against a current condition. A project plan may contain broader work, but it should not replace a clear baseline and experiment.

Can AI identify the root cause?

It can organize evidence and propose hypotheses, but it cannot verify a cause without observations or data. Test hypotheses before committing to a fix.

How many actions should a kaizen include?

Use only the few actions needed to test the first countermeasure. If the list requires multiple teams or months of work, split it into smaller kaizens.

Related prompts

All Operations & Project Management prompts →

Get a few prompts like this every week

Something Big is a free AI newsletter read by 50,000+ professionals. One email a week with prompts and tools that work, plus what changed in AI and what to do about it.