A3ISO 216 A3 paper size (297 × 420 mm) — Toyota one-page structured problem-solving artefact

An A3 is a single sheet of A3 paper (297 × 420 mm — roughly 11 × 17 inches in US measure) that captures a problem, proposal, or status in a structured visual format. The constraint of one page is the point: it forces the author to prioritise ruthlessly, to think visually rather than write paragraphs, and to integrate background, analysis, countermeasure and follow-up into a single coherent argument that a stakeholder can absorb in five minutes.

The discipline behind the artefact matters more than the paper. An A3 is a manifestation of PDCA (Plan-Do-Check-Act) thinking: left side captures the Plan (background → current condition → target → root-cause analysis → proposed countermeasures), right side captures the Do-Check-Act (implementation plan → results → follow-up and standardisation). The structure forces the author to do the thinking — to make their reasoning visible, to expose assumptions, and to commit to verifiable outcomes. Toyota's experience is that the act of authoring an A3 develops the author's problem-solving capability; the document is the by-product.

Toyota uses four distinct A3 templates. Each has a different purpose, audience, and shape — confusing them is a common early mistake.

The problem-solving A3 is the most common and the most rigorous. It follows the canonical 7-box layout: title + owner + date (header), background (1), current condition (2), goal/target (3), root-cause analysis (4), countermeasures (5), implementation plan (6), follow-up / results (7). The proposal A3 looks similar but emphasises the rationale and the financial / strategic case. The status A3 mirrors the original problem-solving A3 with actual results overlaid on the planned results. The information A3 is the least structured but still constrained to one page.

Read left-to-right, top-to-bottom. Every box is essential; skipping any box weakens the argument and almost guarantees the countermeasure will fail.

1. Title + owner + date + version. The title is a one-line statement of the problem or theme — not a topic. 'Reduce line 3 changeover time from 47 min to 15 min by Q3' beats 'Line 3 efficiency.' Owner is a single name (accountability cannot be shared). Date is the last revision date; version is incremented every substantive change.
2. Background. Why does this problem matter now? What is the strategic context, the customer impact, the regulatory pressure? Two or three sentences plus a small graphic. Stakeholders who do not understand the 'why' will not engage with the rest.
3. Current condition. The factual state at the gemba — measured, not described. Use process maps, run charts, Pareto charts, or photos. This is where genchi genbutsu is visible: the author has gone and seen, and is showing the reader what they saw. Avoid opinions; data only.
4. Goal / target condition. The specific, measurable, time-bound state the project will achieve. 'Improve quality' is not a target; 'reduce line 3 first-pass defect rate from 4.1% to <1.0% by 30 June' is. The gap between current and target is the explicit problem to be solved.
5. Root-cause analysis. The why behind the gap. 5 Whys, fishbone diagram (4M + E — Man, Machine, Method, Material, Environment), Pareto, or a more sophisticated technique. The constraint of one page forces the author to surface only the validated root causes — speculation gets cut.
6. Countermeasures. The specific actions that address the validated root causes. Each countermeasure traces back to a root cause; orphan countermeasures (actions without root-cause justification) are a red flag. Include expected effect per countermeasure where possible.
7. Implementation plan + check schedule. Who does what by when, plus when and how the team will check whether each countermeasure worked. The check schedule is what separates an A3 from a wish list — committing to a measurable check date is how PDCA closes the loop.

The A3 is not a deliverable handed to a coach for review. It is the trace of a coaching dialogue. In Toyota's model the author drafts an A3, brings it to their coach (typically their direct manager), and the two have a structured conversation. The coach asks questions — never gives answers — that surface the author's thinking, expose assumptions, and develop the author's problem-solving capability. The author leaves with the A3 marked up, returns to the gemba, refines the analysis, and brings v2 back. A mature A3 cycle produces 6–10 revisions, each one visibly improved.

John Shook describes this in Managing to Learn through the case of Porter (a manager) and Sanderson (his coach). Sanderson's questions are not 'have you considered X?' (which gives the answer) but 'what does the data actually show?' or 'how did you validate that root cause?' or 'what would happen if your countermeasure does not work?' — questions that force Porter to do the thinking. The dialogue is exhausting and slow at first; over months Porter internalises the questions and starts asking them of himself before bringing v1 to Sanderson at all. That internalisation is the real product.

Mike Rother's Toyota Kata generalises this as the 'coaching kata' — a five-question structure (current condition / target condition / obstacles / next step / when can we see what was learned) that any coach can use to guide any author through an improvement. The A3 is the physical artefact the coaching kata revolves around.

A mature A3 is never published in v1 form. It evolves through nemawashi, coaching dialogue and gemba validation. The version stamp is not decorative — it is the auditable trace of the thinking process and the cross-functional alignment. A typical evolution for a substantive problem-solving A3 in regulated manufacturing:

Plants new to A3 work often skip versioning — every conversation overwrites the same v1 with no trace of the evolution. The cost is that nobody can see how the proposal developed, dissenting voices that were addressed look the same as dissenting voices that were ignored, and the coaching dialogue evaporates. Versioning is cheap; not versioning is expensive.

In regulated manufacturing the A3 is not a replacement for the regulated record — it is the thinking substrate that produces a better regulated record faster. The A3 sits underneath change control, CAPA, deviation investigation, kaizen events, hoshin tactics and management review; the formal regulated record (change order, CAPA file, deviation closure, hoshin scorecard, management-review minute) is generated from the A3 once the A3 has converged through nemawashi.

• 21 CFR 820.100 / ISO 13485 §8.5.2 — CAPA. The problem-solving A3 maps directly onto a CAPA investigation: problem statement → current condition → root cause (validated, not assumed) → countermeasure → effectiveness check. Plants that run CAPA-as-A3 see investigation cycle time fall by 30–50% and effectiveness-verification rates rise from typical 60–70% to 90%+ because the root-cause discipline is tighter.
• 21 CFR 211.192 — production record review (deviation discipline). A deviation investigation done as an A3 surfaces the validated root cause within hours of the deviation rather than days, because the structure forces the discipline rather than tolerating long narrative write-ups.
• 21 CFR 211.100 / ISO 13485 §7.3.9 / EU GMP Ch.4 — change control. A change-control proposal authored as a proposal A3 and circulated through nemawashi before the CCB collapses CCB cycle time from 4–6 weeks to 1–2 weeks. The CCB reviews a converged proposal, not a raw idea.
• ICH Q9(R1) — quality risk management. The A3 root-cause section + countermeasure section is where QRM thinking lives operationally. Risk-rating per root cause + risk-reduction per countermeasure is a natural overlay.
• ICH Q10 §3.2.4 — continual improvement. Kaizen-event outputs documented as A3s and stored in the PQS become the audit-evidence trail for §3.2.4 — 'evidence of continual improvement of process performance and product quality.'
• ICH Q10 §2.7 / 21 CFR 820.20 / ISO 13485 §5.6 — management review. Each management-review agenda item presented as a one-page A3 (rather than a 30-slide deck) compresses the review and forces executives to engage with the actual data rather than the curated narrative.
• Hoshin kanri catchball — each catchball round produces a refined X-matrix (A3-equivalent at strategy scale); each tactic at the team level is owned by a problem-solving A3 with its own coaching dialogue.

• A3 cycle time — from first draft to signed decision. Healthy: 2–4 weeks for a substantive problem-solving A3. <1 week means insufficient nemawashi; >8 weeks means the problem is unclear or scope is wrong.
• Revision count per A3 — typically 4–8 revisions for a healthy substantive A3. <2 means coaching dialogue is performative; >12 means scope is unclear or the author is iterating without convergence.
• Coach engagement rate — % of A3s with logged coaching dialogue (not just review). Toyota cultures: 95%+. Western copy-paste-A3 cultures: <30% — coaches sign off rather than coach.
• Effectiveness-verification rate — % of A3s where the planned check confirms the countermeasure achieved the target. World-class: 80%+. Below 60% means the root-cause analysis is weak.
• Repeat-occurrence rate — % of problems where the same A3 has to be reopened within 12 months. World-class: <5%. Above 20% means countermeasures address symptoms not causes.
• Author development trajectory — over 12–18 months, individual authors should need fewer coaching cycles, write tighter root-cause sections, and propose more targeted countermeasures. The author's A3 quality curve is the real outcome.
• A3-to-formal-record cycle time — for CAPA, change control and deviation, time from A3 v1 to signed regulated record. Plants with mature A3 discipline halve this metric and raise audit findings on rigour simultaneously.

1. Jumping straight to countermeasures. Boxes 2–4 (current condition / target / root cause) get filled in after the countermeasure is chosen, as decorative justification. The A3 then becomes confirmation bias on paper.
2. Treating the A3 as a deliverable rather than a thinking artefact. Author writes alone, hands to coach for review, coach signs. No dialogue, no development. Two months in, the practice withers.
3. Coach giving answers instead of asking questions. The coach 'helps' by marking up the A3 with their own analysis. The document gets better; the author learns nothing.
4. Skipping versioning. Every revision overwrites the same file. Nobody can trace how the proposal evolved or whose input was accepted vs ignored. Trust evaporates.
5. Stretching beyond one page. Multi-page 'A3s' are status reports. The constraint of one page is what forces prioritisation, visual thinking, and root-cause clarity. A two-page A3 is a failed A3.
6. PowerPoint-instead-of-A3. Decks defeat the discipline — every box becomes its own slide, the integration disappears, and the reader cannot see the whole argument at once. A3 must be a single visual surface.
7. Author writes the A3 from their desk. Genchi genbutsu (go and see) is a prerequisite for box 3 (current condition). A3s authored from the office without gemba time produce countermeasures that fail at the gemba.

V5 ships A3 as the thinking substrate under CAPA, change control, deviation investigation, kaizen events, hoshin tactics and management review — and as the coaching workspace that develops problem-solvers, not just solves problems. The platform enforces the discipline (one page, structured boxes, version stamping, coaching dialogue) while integrating the A3 with the regulated record so the formal CAPA / change-control / deviation file is generated from the A3 rather than authored separately.

• Four A3 templates — problem-solving (7-box), proposal, status, information. Each template enforces the canonical structure with required boxes; the author cannot publish v1 with empty mandatory boxes.
• Single-page constraint. The authoring canvas is fixed at A3 proportions; the platform refuses to add a 'second page.' Authors learn to prioritise.
• Visual blocks. Each box accepts text, a chart (run chart / Pareto / control chart / fishbone), a process map, a photo from the gemba, or a small table. Inline editing; no embedded PowerPoint.
• Version stamping with delta summary. Every revision auto-versions; the author writes a one-line delta summary describing what changed and why; previous versions remain immutable and viewable side-by-side.
• Coaching dialogue capture. Each coaching interaction (in-person, video, chat) is logged against the A3 with date, coach, questions asked (free-form or kata-five-question template), author responses, and resulting next step. Becomes the audit trail of the dialogue + the personal-development record for the author.
• Toyota Kata five-question template. Optional structured coaching mode — current condition / target condition / obstacles / next step / when can we see what was learned — captured per coaching session.
• Integrated nemawashi workflow. The A3 plugs directly into the nemawashi engine — stakeholder matrix, conversation log, readiness gate, pre-meeting alignment — so the A3 is the artefact that travels through every nemawashi conversation.
• Regulated-record generation. When the A3 converges and is signed, V5 generates the linked CAPA file / change-control order / deviation closure / kaizen-event record automatically from the A3 content. One source, two views — the A3 (for thinking) and the regulated record (for auditing).
• Effectiveness verification scheduling. Each A3 has a mandatory check date; V5 auto-creates the verification task on the date, links it back to the A3, and tracks whether the countermeasure achieved its target. Feeds the effectiveness-verification rate metric.
• Obeya integration. A3s under nemawashi auto-publish to the obeya wall as cards; status A3s auto-update from live data; the wall becomes a real-time view of the programme's thinking, not a snapshot.
• Author quality curve. Per-author dashboard tracks A3 cycle time, revision count, coaching engagement, effectiveness-verification rate, and repeat-occurrence rate over time — the author's development trajectory becomes visible to them and their coach.
• Regulated overlay (820.100 / 211.192 / 211.100 / 820.30(i) / 820.20 / ISO 13485 §8.5.2 / §7.3.9 / §5.6 / ICH Q9(R1) / ICH Q10 §2.7 / §3.2.4 / EU GMP Ch.4). The same A3 substrate serves CAPA, change control, deviation, management review and continual improvement — auditors see one coherent thinking discipline, not parallel narratives.
• Part 11 + Annex 11 audit trail. Every A3 revision, coaching dialogue log, stakeholder position update, version sign-off and effectiveness-check result is timestamped, attributed, immutable, and retained.
• Mobile-safe (iPhone ≤390px). A3s are read at the gemba; the layout reflows for mobile without losing the single-page mental model.