An A3 problem-solving conversation should help the problem owner explain the gap, examine evidence, test explanations and plan follow-up. The coach's task is not to fill in a polished page on the person's behalf. A well-formatted sheet can hide weak reasoning as easily as reveal clear reasoning.
The Lean Enterprise Institute's A3 resource presents A3 as a problem-solving and management discipline, not merely a paper size. These questions are a general instructional aid. They do not replace technical expertise, approved investigation methods or required safety and quality procedures.
Background: why does this problem matter?
Ask which customer or operating need is affected and why this problem deserves attention now. What responsibility does the problem owner hold? What is inside and outside the scope? A proposed improvement that cannot name the affected work may be an interesting idea rather than a priority.
Avoid forcing every problem into a grand business objective. A bounded recurring delay can be worth examining on its own. The background should provide enough context to understand the choice, not an impressive story assembled after a solution has already been selected.
Current condition: what do we actually know?
Ask what was directly observed, what was reported by others and what remains an interpretation. Where in the workflow does the gap appear? How often was it seen, in what sample and under which conditions? What approved record or observation supports the description?
Use gemba questions when appropriate to understand the work with the people doing it. If the evidence is unavailable, make gathering it the next step. Do not reward a confident explanation that outruns what the team knows.
Target condition: what would better look like?
Define the expected condition and a meaningful measure. Ask what should change, what must remain protected and how the team would recognize the difference. A target should relate to the gap, not simply repeat a general request to work harder or communicate better.
For a fictional handoff problem, a target might be that transferred work has an accepted next owner before it enters the receiving queue. That is more observable than seamless collaboration. Required authorizations and operating standards still govern what the team may change.
Analysis: which explanation is supported?
Ask the problem owner to explain the link between an observed condition and the proposed cause. What evidence would challenge that explanation? Did the team examine a comparable situation where the problem did not occur? What alternative explanation remains plausible?
Do not turn repeated why questions into a demand that an employee confess a root cause. A lack of effort is not established by a missed target. The investigation might reveal information timing, authority, conflicting standards or another condition. The coach helps test reasoning rather than select a convenient person to blame.
Countermeasures: why should this change address the gap?
Ask which part of the supported explanation the proposed action addresses. What alternatives were considered? What is the smallest authorized test that could provide useful evidence? Who needs to be involved before the test begins?
A new form may help expose a missing owner, but it cannot assign authority that leadership has not clarified. Make that dependency visible. Identify boundaries and approval requirements so the learning experiment does not become unauthorized process change.
Implementation: who owns the work and support?
Capture the next action, responsible person, needed input and review point. Distinguish the person doing the investigation from the person authorized to approve the operating change. Ask the receiving team what it needs to participate meaningfully.
The manager coaching template helps keep ownership with the learner. If the coach takes over every uncertain step, the A3 may be finished while the person's problem-solving capability remains unchanged.
Follow-up: what happened compared with the prediction?
Review what the team expected and what the test actually showed. Did the condition improve under comparable circumstances? Did another problem appear? Was the action applied as intended? Keep the definition of the measure consistent so a favorable result is not created by changing the count.
Choose whether to retain, adjust or abandon the countermeasure using that evidence. An A3 is useful when it supports better thinking and follow-through, not when every section is filled. For problems recurring across leadership levels, discuss the operating situation rather than asking for a more attractive template.
About this resource: Prepared with AI-assisted drafting from published guidance. Examples are illustrative, not reported client results. Cover illustration: AI-generated.