Many improvement efforts fail in the same way. Someone notices a problem, a meeting jumps straight to a solution, the solution is implemented, and a few months later the problem is back or has moved somewhere else. Nobody checked whether the solution addressed the real cause, nobody measured the result and nobody wrote down what was learned. The next team repeats the cycle.
The A3 method, developed within Toyota and now used widely in manufacturing, healthcare, services and offices, is a simple antidote. It takes its name from the A3 sheet of paper, 297 by 420 millimetres, on which the whole problem-solving story is written: the background, the current condition, the target, the root causes, the countermeasures, the plan and the results. The single page forces clear thinking and concise evidence, makes the reasoning visible to others, and follows the plan-do-check-act cycle that underpins disciplined improvement.
The deeper purpose of A3 thinking is to develop people. In organisations that use it well, managers coach staff through A3s, asking questions rather than giving answers, so that people closest to the work learn to understand problems and improve their own processes. This article explains the PDCA cycle, the structure of an A3, how to complete each section well, how managers use A3s for coaching, other uses of the format and common pitfalls. It is general information for supervisors, team leaders, managers and improvement specialists.
Plan-do-check-act
Plan-do-check-act, often called PDCA or the Deming cycle after the quality pioneer W. Edwards Deming, who popularised it from Walter Shewhart’s earlier work, is a cycle for learning and improvement:
- Plan: understand the problem and its causes, set a target and plan countermeasures.
- Do: carry out the countermeasures, often first as a trial.
- Check: measure the results against the target and study what happened.
- Act: standardise what worked, adjust what did not and decide the next cycle.
Most of the effort belongs in planning, particularly in understanding the current condition and causes. Teams that rush planning tend to cycle through solutions that do not work. The cycle repeats: each turn builds on what the last one taught. Structured methods such as DMAIC follow the same logic at larger scale; the fixing a recurring problem with DMAIC article covers that approach.
The structure of an A3
A typical problem-solving A3 has two halves. The left side tells the story of the problem; the right side tells the story of the solution.
| Section | Question it answers |
|---|---|
| Title and owner | What is this about, and who is responsible? |
| Background | Why does this problem matter to the business and customers? |
| Current condition | What is actually happening, shown with facts and data? |
| Target | What specific, measurable result do we want, by when? |
| Analysis | Why is this happening, down to root causes? |
| Countermeasures | What will we change to address each root cause? |
| Plan | Who will do what, by when? |
| Check | What happened, compared with the target? |
| Follow-up | What will we standardise, share and do next? |
Formats vary, and that is fine. The thinking matters more than the template.
Completing each section well
Background
Explain briefly why the problem matters: its effect on safety, quality, delivery, cost, customers or staff, and how it connects to the organisation’s goals. A reader should understand why this problem deserves attention rather than others.
Current condition
This is the heart of the A3 and the most often skimped. Go and see the work where it happens, sometimes called going to the gemba, rather than relying on reports. Describe the process as it actually runs, with a simple diagram, and quantify the problem: how often, how much, where, when and for whom. Separate data by source to find where the problem concentrates. A Pareto chart of causes or categories often belongs here.
A good current condition statement is factual and specific, such as “Picking errors averaged 1.2% of order lines over the last 12 weeks, about 600 errors a month, and 45% involved look-alike products stored next to each other”, not “Picking accuracy is poor”.
Target
State a measurable target and a date, such as “Reduce picking errors to 0.4% of lines within three months, without slowing picking rates.” The target should be challenging but grounded in the analysis of what can be changed.
Analysis
Find root causes, not symptoms. Common tools include the five whys, asking why repeatedly until reaching a cause that, if removed, would prevent recurrence, and cause-and-effect diagrams. Test suspected causes with data or simple trials where possible. A root cause is usually a condition in the process, such as a layout, a method, missing information or an unclear standard, rather than an individual’s carelessness.
Countermeasures
The word countermeasure is used deliberately instead of “solution”, recognising that each change is a hypothesis to be tested, not a permanent answer. Link each countermeasure to the root cause it addresses. Prefer changes that make errors hard to make, such as physical separation or error-proofing, over reminders and retraining alone. Where several countermeasures are possible, compare them briefly on effect, cost and speed, and consider trialling the most promising on a small scale before rolling it out.
Plan
List actions with owners and dates. Keep it short and realistic.
Check
After implementation, measure the result using the same measure as the current condition, compare it with the target and explain any gap. Unexpected effects, good or bad, belong here too.
Follow-up
If the countermeasures worked, update standard work, training and visual controls so the improvement holds, and share the lessons with other areas that could benefit. If they did not fully work, plan the next cycle.
A3 as a coaching tool
In organisations where A3s work well, they are not reports written for managers to approve. They are a means for managers to coach. The owner drafts the A3; the coach, usually their manager or a more experienced colleague, asks questions:
- How do you know that? Have you seen it yourself?
- What does the data show when you separate it by shift, product or location?
- Why does that happen? And why does that happen?
- How does this countermeasure address that cause?
- How will you know whether it worked?
- Who else needs to agree, and who else could benefit?
The coach resists the temptation to supply answers. Over time, the owner learns to see problems clearly and think them through, which builds the organisation’s capability to improve. Experienced lean practitioners often describe the goal as developing people who improve the system, rather than having specialists improve it for them. The lean as an operating system, not a toolkit article discusses this wider view of improvement.
Other uses of the A3 format
The same one-page discipline suits other purposes:
- Proposal A3s set out a recommendation, such as a new piece of equipment or a change of process, with background, current condition, options analysis, recommendation, costs and benefits, and plan.
- Status A3s report progress on a project or improvement: original plan, results so far, remaining issues and next steps.
- Strategy A3s summarise a team’s goals, the gap to be closed and the main initiatives for a year.
A3 proposals are often shared widely in draft, gathering input and building agreement before a formal decision, a practice sometimes called nemawashi. By the time the decision is made, most concerns have already been addressed.
A3s beyond the factory
The method works in any setting where work can be observed and measured. A clinic can use an A3 to reduce patient waiting times, mapping the patient journey and finding where delays build up. An accounts team can tackle late supplier payments by examining where invoices stall. A sales team can investigate why quotes take too long. An engineering office can study why drawings are frequently revised after issue. In each case, the discipline is the same: see the work, measure the problem, find causes, test countermeasures, check results and standardise.
Getting started
Organisations new to A3 thinking can start small:
- Train a few leaders in the method and in coaching through questions, using real problems from their own areas.
- Pair each A3 owner with a coach, and hold short, regular coaching conversations, often weekly, rather than occasional long reviews.
- Start with modest, local problems that can be solved within a few weeks, so people experience a full cycle and see results.
- Keep a library of completed A3s, organised by area and problem type, so others can learn from them.
- Recognise good thinking, not just good results: an A3 that reveals a problem is bigger than expected, or that a countermeasure failed, still adds knowledge.
- Expand gradually as coaches develop, because the coaching capacity, not the template, limits how fast the method spreads.
Leaders set the tone. When senior managers write their own A3s, ask questions at the workplace and respond to problems with curiosity rather than blame, others follow.
Keeping A3s practical
- Choose the right size of problem: one that the owner can understand and influence within weeks or a few months. Very large problems need to be broken down.
- Write by hand or simply: elaborate formatting wastes time; clear diagrams, honest data and concise words matter far more.
- Limit the number of A3s under way at once, so each gets proper attention and coaching.
- Display A3s in the work area or on the team board, so progress is visible to everyone involved.
- Review completed A3s to share learning across teams and sites.
Frequent small improvements, each with a simple PDCA cycle, often add up to more than occasional large projects. The small, frequent improvement on the shop floor article covers building that habit.
A worked example
This is an illustrative example. A distribution centre picks about 50,000 order lines a month. Customer complaints about wrong items are rising, and the operations manager asks a team leader to work through an A3, with coaching.
Current condition. The team leader spends time on the floor and reviews 12 weeks of error records. Errors average 1.2% of lines, about 600 a month. A Pareto chart shows 45% involve look-alike products stored next to each other, 25% involve unclear or damaged location labels, 15% are made by staff in their first month and the rest are spread across other causes. Errors are higher on the evening shift, when fewer experienced pickers work.
Target. Reduce errors to 0.4% of lines within three months.
Analysis. Asking why repeatedly, the team finds that look-alike products end up adjacent because new products are slotted wherever space is free, with no rule about similarity. Labels are damaged by forklift contact at low levels. New starters receive a half-day induction and then work alone.
Countermeasures.
- A slotting rule separates look-alike products, and existing look-alike pairs are moved apart.
- Durable labels are fitted in protected positions at low levels.
- Scanning confirmation is added at pick for the 200 highest-risk products.
- New starters work alongside an experienced picker for their first two weeks.
Check. After eight weeks, errors average 0.45% of lines, about 225 a month, close to but above the target. The remaining errors are analysed; most involve a product family with similar packaging from one supplier. A second PDCA cycle adds that family to scanning confirmation, and errors fall below 0.4%.
Follow-up. The slotting rule and labelling standard are written into standard work, the new starter buddy system becomes standard, and the A3 is shared with the company’s other distribution centre.
Applying this in an Australian business
- Use PDCA as the basic rhythm for improvement.
- Spend most effort on the current condition, with facts gathered at the workplace.
- Set measurable targets with dates.
- Find root causes and test them.
- Treat countermeasures as experiments, and check results honestly.
- Standardise and share what works.
- Coach with questions, so people learn to solve problems.
- Use A3s for proposals and status reports as well as problems.
Where A3s go wrong
- Filling in a form instead of thinking.
- Jumping to solutions before understanding the current condition.
- Vague problem statements without data.
- Root causes that blame individuals.
- No check step, so nobody knows whether anything worked.
- Managers writing or approving A3s instead of coaching.
- Too many A3s at once.
- Completed A3s filed away where nobody else can learn from them.
Questions to ask about an improvement
- What exactly is the problem, and how do we know?
- Have we seen the work ourselves?
- What is the target, and by when?
- What are the root causes, and how were they confirmed?
- How will we know if our countermeasures worked?
- What will we standardise and share?
Bringing it together
The A3 method puts a complete problem-solving story on one page and follows the plan-do-check-act cycle: understand the current condition with facts, set a target, find root causes, test countermeasures, check the results and standardise what works. Its greatest value lies in coaching, as managers ask questions that help people closest to the work learn to solve problems themselves. Used for problems, proposals and status reports, with a manageable number under way at once, A3s build an organisation that improves continually and remembers what it has learned.
Source: KEVOS editorial notes, drawing on earlier KEVOS operations excellence study material on lean practice, the Toyota Production System and continuous improvement, together with established improvement practice. The worked example is illustrative. This article is general information.
