A3 problem solving

What it is

A3 emerged at Toyota alongside the Toyota Production System through the mid-20th century; Sobek and Smalley’s Understanding A3 Thinking (Productivity Press, 2008) is the canonical English-language treatment. The document is seven sections arranged on a single page — the left side characterizes the problem and establishes the root cause; the right side proposes countermeasures and tracks whether they worked. A3s were the currency of the Toyota Production System — engineers, managers, and executives wrote them, shared them, and updated them as problems progressed. The seven-section sequence used here follows Sobek and Smalley’s formalization.

The method’s real power is its cadence. An A3 isn’t written once and filed. It evolves — the initial draft characterizes the problem; after investigation, the root cause section gets filled in; after implementation, the follow-up section tracks whether the metric actually moved. The A3 becomes a living artifact that carries understanding from one sprint to the next.

For our purposes in waste-walk workshops, an A3 is the right tool when a root cause you’ve surfaced is too big to resolve in a single sprint — it needs sustained attention, possibly across multiple teams, possibly with budget or org-change implications. A sticky-note capture loses context over time. An A3 doesn’t.

When to use it

Produce an A3 when the root cause meets at least one of these criteria:

• The fix will require more than one sprint’s action to close.
• It spans multiple teams or contexts (not just the team running the waste walk).
• It’s been observed across 2+ consecutive waste walks — a recurring pattern, not a one-off.
• It will benefit from leadership visibility — a budget ask, an external dependency, an org-change conversation.

Don’t produce an A3 when:

• A single sprint’s action will fully resolve the problem. A regular retrospective action item is enough.
• The root cause is trivial and the action is obvious.
• The team doesn’t have the time or discipline to maintain the A3 across multiple reviews. A stale A3 is worse than no A3 — it creates an illusion of progress without any.

A3 is explicitly called out in the rework workshop as the escalation path for systemic rework causes that span multiple sprints or teams. It works equally well for recurring automation debt and planning dysfunction surfaced in the other two workshops.

The seven sections

Fill these in order. The countermeasures and plan (items 5, 6 and 7) are blocked by the problem characterization and root cause (the first 4 items). Don’t skip ahead.

Blocking side — understand the problem:

1. Background. Why does this matter? Business context, impact, why now. Two to three sentences. If you can’t explain why this A3 is worth writing in three sentences, the problem isn’t clear enough yet.
2. Current Condition. What’s actually happening today? Facts, measurements, observed behavior. This is where the waste-walk numbers live — “rework hours in origin category X: 12 this sprint, 10 prior, 6 prior. Trend: growing.”
3. Goal / Target Condition. What does “fixed” look like? Specific, measurable, time-bound. “Reduce Upstream-category rework hours from 12 to ≤3 per sprint, by end of Q2.” If you can’t name a metric, you don’t have a goal — you have a wish.
4. Root Cause Analysis. The 5 Whys chain or fishbone that led you to a terminal root cause. Include only the chain that actually reached the root; strip the exploratory branches that didn’t.

Action side — act on it:

5. Countermeasures. Specific changes that address the root cause. Not “we’ll try harder” — concrete interventions. Multiple are fine; each should be numbered.
6. Implementation Plan. Owner, steps, schedule. Each step has a date and a done-definition. Without these, the A3 stalls.
7. Follow-up / Check Results. How will you know the countermeasures worked? What metric will you watch? When’s the review? Best practice is linking the A3 back to a waste walk so the next sprint’s readout is the verdict.

The template

A traditional A3 fits on a single page. In the canonical Toyota layout, sections 1–4 occupy the left half of the page (problem characterization) and sections 5–7 occupy the right half (countermeasures and follow-through). When you fill one in digitally, the same seven sections work just as well stacked.

Above the sections, a single metadata line identifies the A3: title · owner · sponsor · date drafted · review date.

Worked example

This is the A3 produced from the rework workshop’s worked example, after 5 Whys surfaced the root cause.

Notice what the A3 does that a sticky note can’t. It carries the reason for the intervention alongside the intervention itself — the chain of whys is right there on the page. Six months from now when a new PO asks “why do we do this mid-sprint review?”, the A3 is the answer. And the follow-up section means the A3 gets re-read — it doesn’t disappear into a wiki graveyard.

Common failure modes

• Skipping Current Condition straight to Countermeasures. Symptom: the countermeasures aren’t tied to facts, they’re tied to opinions. The team’s fastest talker is now setting direction. Fix: go back and quantify the current condition. If you can’t, you’re not ready to act.
• Vague root cause. Symptom: the root cause reads like “we need better communication” or “we should be more careful.” These aren’t testable, and countermeasures against them will be equally vague. Fix: keep asking why until you reach something specific — an absent ritual, a missing owner, a broken cadence.
• Countermeasures without owners or dates. Symptom: the A3 sits. Nobody owns the work, so the work doesn’t happen. Fix: every row of the Implementation Plan table has a person’s name and a date before the A3 is considered drafted.
• A3 becomes a full essay. Symptom: spills over one page. The author is using the document to work out their own thinking instead of communicating conclusions. Fix: the page limit is the method. Cut until it fits. If it genuinely won’t fit on one page, you probably have two problems, not one — split the A3.
• The A3 is drafted and never re-opened. Symptom: countermeasures got implemented (or didn’t), but the follow-up section never gets filled in. The metric moves (or doesn’t) and nobody notices. Fix: schedule the review date on the calendar when the A3 is drafted, not as a hopeful intention. Review it. Close it when the criteria are met. Re-open it if the metric regresses.

The discipline around A3

A3 is often described as “lean problem solving,” but that undersells what it actually is. It’s a communication protocol. John Shook’s Managing to Learn (LEI, 2008) frames it as the way Toyota managers taught problem solving — by walking a subordinate through an A3 draft, asking probing questions, and sending them back to investigate the gaps. The document is the output, but the process of producing it is the real product.

For a distributed team doing waste walks, you won’t have a Toyota-style sensei sitting next to you. The substitute is review discipline: every A3 gets a second reader before it’s considered drafted, and every A3 gets revisited at its review date. Without those two rituals, A3s decay into write-once documents and the method stops working.

References

In the playbook

• Rework workshop — A3 is introduced in the “Alternative and complementary exercises” section as the escalation path for systemic root causes.
• 5 Whys — the root-cause chain that populates Section 4 of an A3.
• Fishbone diagram — the multi-factor alternative to 5 Whys, equally valid for populating the Root Cause Analysis section.

External references

• Durward K. Sobek II and Art Smalley, Understanding A3 Thinking: A Critical Component of Toyota’s PDCA Management System (Productivity Press, 2008) — the canonical English-language treatment. This is the book if you only buy one on A3.
• John Shook, Managing to Learn: Using the A3 Management Process to Solve Problems, Gain Agreement, Mentor, and Lead (Lean Enterprise Institute, 2008) — the companion guide, told as a narrative of a manager coaching a subordinate through their first A3.
• Lean Enterprise Institute, A3 Report — online reference with worked examples.
• Lean Enterprise Institute, Templates and worked examples — the LEI’s template library includes A3 variants for different problem types.
• Jeffrey Liker, The Toyota Way (McGraw-Hill, 2004) — broader context for where A3 sits within the Toyota Production System.