Operations Deliver

A3 Problem Solving

A3 problem solving is one sheet that holds the problem, the analysis, the countermeasures, and the follow-up, written to be discussed and revised.

The questionWhat is the problem, what causes it, what will we try, and how will we know it worked?

A3 report. Named for ISO 216 A3 paper, 297 by 420 millimeters, close to 11-by-17-inch tabloid.

Process · Toyota, written up by Durward Sobek and Art Smalley, Toyota practice, 2008 in English

Background Why this problem, why now
Current condition Facts about how it runs today
Goal The target and how we will know
Analysis Causes, tested, not a symptom list
Countermeasures What we will try, and why
Plan Who, when, and what
Follow-up How we check and what we standardize

Typical A3 sections. The first four describe the problem. The last three describe the response.

When to use it

A problem has been talked about for months, and there is still no one-page statement of the current condition, the cause, and the trial.
A manager wants to coach someone through problem solving rather than hand them the answer.
Several functions disagree about the facts, and a shared sheet would force the argument onto evidence.
Hoshin or daily management has thrown up a gap that needs structured PDCA, not a task list.

When to avoid it

The cause is already known and the work is implementation. Write the plan. Do not dress it as an A3.
The problem is a portfolio, a strategy choice, or a political conflict. One sheet cannot hold it.
Nobody will walk the work or collect facts. An A3 filled from opinion is a slide with boxes.
Leadership wants the template completed by Friday as proof of lean. The coaching cycle takes longer than that.

Overview

An A3 is one sheet that holds the problem, the facts of the current condition, the cause, the trial, and the follow-up, written so it can be discussed and revised. It supports a coaching choice: whether to hand someone the answer, or to make them show the reasoning on a page that other people can challenge.

The layout is not the method. Sheets filled backward from a preferred fix, or closed when the plan is written, are status reports on A3 paper. The working product is a current condition with numbers, a cause that would change those numbers, and a follow-up that checks the goal. Expect several drafts.

The typical sections

Background

Why this problem, why it matters, and why now. Enough context that a reader who does not run the process can follow. Keep it short. Background that retells the company history is padding.

Current condition

How the work runs today, with facts and a picture. Counts, times, locations, and who is affected. A current-condition box with no number and no sketch is not done.

Goal

The target condition, the measure, and the date. It should be checkable. "Improve quality" is not a goal. "Cut mis-picks in zone A from 1.8 percent to below 0.8 percent by 15 March" is.

Analysis

The causes that explain the current condition, tested by going to see and by data. Fishbone and 5 Whys are common tools here. Analysis that restates the symptom, or lists ten untested causes, is not analysis.

Countermeasures

The changes that attack those causes, each linked to a cause. Several small trials beat one large initiative. If a countermeasure does not map to a cause on the page, it is a pet idea.

Plan

Who will do what, by when, and what they will look at. A Gantt is optional. Named people and dates are not.

Follow-up

How the result will be checked against the goal, what will be standardized if it holds, and when the sheet will be reviewed. This is the Act of PDCA. Sheets that stop at the plan never close.

What it is not

A template. Filling boxes in order is not A3 thinking. The point is the conversation that changes what is on the page.
A status slide. Red-yellow-green on a task list is a report. An A3 is supposed to show the problem and the reasoning.
A substitute for going to see. If the current condition was written at a desk, the analysis will be too.

The steps

Name an owner and a problem that fits one sheet

One person owns the A3. The problem should be small enough that current condition can be drawn and counted on the page. If it needs a program office, it is too big.
Write current condition from the work

Go to the process, count, sketch, and date the facts. Write background and a checkable goal only after the current condition is on the page. Output is the left side as facts, not as a complaint.
Analyze until a cause can be shown

Use a fishbone, 5 Whys, or stratified counts. Keep asking until a cause would, if removed, change the current-condition number. Cross out causes you cannot support.
Propose countermeasures and a plan, then discuss

Link each countermeasure to a remaining cause. Put owners and dates on the plan. Walk the sheet with the people affected and with the coach. Expect the page to change. The discussion is the method.
Trial, check, and standardize

Run the plan, write what happened in follow-up, and either adopt a new standard or revise the analysis. Do not close the A3 on the plan date. Close it when the goal has been checked.

In practice

Wrong-item picks in a distribution center sit at 1.8 percent of lines. The operations manager has a scanner proposal and a training plan. The consultant asks the shift lead to put the problem on one A3, starting with current condition, not with the scanners. A week of sample errors, 62 of them, shows 41 clustered among look-alike SKUs stored in adjacent bins in one zone. The other 21 are scattered. The goal is written as 0.8 percent in that zone within six weeks, not a site-wide scanner rollout.

Analysis, using 5 Whys on the 41, stops at location logic that shelves by supplier code rather than by appearance. Countermeasures are a slotting rule for look-alikes, a two-week trial on one aisle, and a confirm-scan only in that aisle. The plan names the slotting owner and the review date. Follow-up after the trial shows that zone's rate at 0.9 percent. The scanner purchase is kept off the sheet until the remaining gap, if any, is measured. The A3 is on its third draft. That is the working product.

What you leave with

One owner and one problem statement that fits the sheet.
A current condition with facts and a sketch, and a checkable goal.
An analysis that names causes and drops the ones that did not hold.
Countermeasures linked to those causes, with a dated plan.
A follow-up record, and an updated standard for anything that worked.

Failure points

Starting from the countermeasure

The sheet is written backward from a preferred fix. Current condition and analysis are filled in to match. The discussion cannot change the answer because the answer was first.

A first draft treated as final

Toyota-style use is revision through conversation. A sheet that is "done" after one sitting is a report, and the thinking never met the people who run the work.

Goal without a measure

"Reduce defects" cannot be checked, so follow-up becomes a narrative of activity. Write the number, the place, and the date.

Analysis as a pile of causes

Every category on the fishbone is copied onto the A3. Countermeasures then multiply to match. Force a few causes that explain the current-condition data.

No follow-up

The plan is presented, the sheet is filed, and nobody compares the result with the goal. PDCA without Check is a task list.

The failures above are usually Synthesis & Story: the work was done, but it never became one clear message a client could act on.

Origin

A3 is a Toyota practice of putting a problem, the thinking about it, and the proposed response on one sheet of A3 paper, then using that sheet as the basis for discussion and coaching. The name is the paper size, not a branded method. Sections vary by company and by whether the sheet is a problem-solving story, a proposal, or a status review. The pattern most consultants mean is PDCA on a page: background and current condition, a goal, cause analysis, countermeasures, a plan, and follow-up.

Durward Sobek and Art Smalley set this out for English readers in Understanding A3 Thinking in 2008. John Shook's Managing to Learn, published the same year by the Lean Enterprise Institute, treats the A3 as a management conversation, not a form. Nobody owns the term. Templates sold under the name are just layouts. The sheet is not licensed. Copying a publisher's filled example from a book is a copyright issue. This note describes the sections in its own words.