The meeting is faster and leaves fewer traces: everyone walks out with their own version, and the cause never gets written down.
A professional working with outside collaborators, a twelve-person company split between office and shop floor, a hundred-employee company with three functions: a problem that passes through several hands behaves the same way — and for the professional, those hands are the client, the collaborator and the supplier.
The arbitration almost always falls to the business owner: among Italian companies under personal or family control with at least 3 employees, only 1.4% rely on an internal or external manager [5] (ISTAT permanent census, reference year 2022).
The A3 report is the format born at Toyota that keeps an entire problem — context, data, goal, cause, countermeasures, plan, verification, standardization — on a single sheet [1].
An A3 is not a summary of a problem: it is the constraint that stops you from jumping to countermeasures before the cause is written down, and it is the document the departments involved read in the same version.
This article shows when to open one, what each of the eight sections pins down, a filled-in case and the mistakes that hollow it out.
When the meeting ends and nothing is left: what Toyota's A3 report is
A problem that was fixed, and a problem that was fixed with the cause written down: six months later, which of the two is still closed?
The difference is not a matter of opinion: the Bank of Italy's Invind survey measures Italian companies on exactly this scale, distinguishing those that simply fix the problem from those that fix it and act to keep it from coming back [4].
A problem that cuts across several departments almost always gets closed the same way — a meeting, an email thread, a verbal agreement — and in the end everyone goes back to their desk with their own version of the facts.
The A3 report was born at Toyota to do the opposite: keep the whole problem on one side of a sheet, in the order it should be tackled, so that people who were not at the table can read it and understand it [1].
The name comes from the paper, not the method: A3 is the 297 × 420 millimeter paper size, at the time the largest a fax machine could transmit [2].
This section defines the tool, distinguishes it from the documents that resemble it and explains why the constraint of fitting on a single side is the part that matters.
That scale is measured on about 3,200 manufacturing and service companies with at least 20 employees, 2019 wave, and classifies what happens when a problem emerges in the production process: from doing nothing to having a process that anticipates similar problems [4].
The details of the eight sections come further on; the short definition is in the glossary.
Meeting minutes record what was said; the A3 records what was verified and what happens on the set date.
PDCA is the logical cycle, the A3 is the sheet that makes it readable to people who were not at the table, and repeating it over time is the subject of continuous improvement.
A technical report has no length limit; the A3 does: the single sheet is the proof that the problem has been understood.
A nonconformity report records the deviation and the corrective action within the quality system (nonconformity management); the A3 is a format for analyzing a problem that is not necessarily a nonconformity.
Six months later, the problem with the written cause is the one still closed, because someone can recheck it on a date — and the step before that is deciding whether it deserves the sheet.
One-hour problems and problems that cut across three departments: when to open an A3
A wrong order sorted out in ten minutes on the phone, and a delay that bounces every month between engineering, production and purchasing: which of the two deserves a sheet?
Only the second, and not because it is more serious: on the first, the sheet costs more than the problem; on the second, it is what stops each department from holding on to its own version.
The sheet is not there to explain the problem to people who weren't present: it is there to stop countermeasures from opening before the cause — which is why on a one-hour problem it is wasted time, and on a problem that cuts across three departments it is what holds the three versions together.
The entry filter comes down to four questions — does the problem recur, does it have a measurable impact in time or money, could it spread to other activities, does doing nothing cost something — and two negative answers are enough to make an A3 overkill.
For companies with a certified quality system, the choice is already partly made: in Italy there were 101,426 active ISO 9001 certificates at the end of 2024 [6], and the standard requires determining the causes of a nonconformity and reviewing the effectiveness of the corrective action [7].
For everyone else it remains a choice of method, and this section gives you the criteria to make it in five minutes instead of after the third meeting.
| Criterion | The question to ask | How it stands here |
|---|---|---|
| Recurrence | Has it already come up in the same quarter? | |
| Economic or quality impact | How many hours, how much money, how many deliveries? | |
| Risk of spreading | Does it affect other jobs, other customers, other departments? | |
| Cost of doing nothing | What happens if it is left as it is for six months? |
With two negative answers there is another route, and it is usually shorter: an isolated anomaly on its first occurrence is closed with a direct response and a recheck date.
A problem within a single function, with a linear causal chain, is dug into with the five whys, while a problem of sequence and idle time is better seen through activity mapping.
A choice among options that are already known is not an analysis problem but a decision problem, and it is weighed with a decision matrix.
The sheet holds up, on the other hand, when the problem cuts across several functions, when the timeline of events is disputed and when the people living the problem are not the ones who decide.
A peer-reviewed case study of a small Italian company documents a process carried out following the A3 model, with data gathered through on-site interviews and countermeasures proposed and adopted; the authors acknowledge the limit of the single case, which describes a process and does not allow outcome estimates to be drawn from it [3].
The upstream step — spotting the problem while it is still a signal — is the subject of business problem finding.
The wrong order sorted out on the phone stays off the sheet, the delay bouncing between three departments goes on it, and the comparison between structured processes is in the guide to business problem solving: the next step is what goes into the eight sections.
The eight sections of the A3 model, one by one: what each pins down and when it is closed
The "current situation" section filled in with "delays have increased," and the same section filled in with "11 jobs out of 24 in the first half": which of the two lets you know, four months from now, whether anything has changed?
Only the second, and the difference is not one of style: without a number in section 2, section 7 has nothing to verify.
The eight sections of the A3 are not chapters of a document; they are mandatory steps in order: background, current situation, goal, cause analysis, countermeasures, implementation plan, results verification, standardization.
Each answers a single question and has a closing criterion, that is, a way to tell it is finished and you can move on.
The Treccani Italian dictionary defines a countermeasure as the measure by which an action is countered or prevented, or its effects neutralized [8]: that is why section 5 bears that name, and the difference is not just terminology — a countermeasure is verified on a date.
This section walks through the eight sections with the question and the closing criterion for each.
| Section | What it pins down | The question | When it is closed |
|---|---|---|---|
| 1. Background | The scope and who is involved | How long has it been going on, which functions does it affect? | When the scope is written down |
| 2. Current situation | The starting data | How much, how often, from which source? | With a number, a period and a source |
| 3. Goal | The measurable target condition | Which number, by when? | When it can be verified against the data in section 2 |
| 4. Cause analysis | The written cause, separate from the hypotheses | Why does it happen, what confirms it? | With one cause verified on the data and the others marked as hypotheses |
| 5. Countermeasures | The actions chosen and those rejected | What counters the cause, at what cost? | When every action has impact, cost and reversibility |
| 6. Implementation plan | Who does what by when | Who is the owner, what is the date? | When every row has a name and a date |
| 7. Results verification | The metric, the threshold and the date | What is rechecked, against what threshold? | When the date is set before closing |
| 8. Standardization | What becomes part of the way of working | Which procedure changes, which recurring check is set? | When the rule is written where the people doing the work can find it |
Section 3 works as a hypothesis, tested in section 7, where you recount the same jobs or the same orders counted in section 2 [2].
Sections 7 and 8 are also what a certified quality system requires: reviewing the effectiveness of the corrective action and updating the procedure that follows from it [7].
In section 4 the techniques have their own pages: the family is root cause analysis, the Ishikawa diagram opens up many disorderly causes, and the 5 whys technique digs down a linear chain.
When there are more than two candidate countermeasures and they need to be weighed, the choice goes onto a decision matrix.
The sequence does not change with the medium: paper, a text document or a spreadsheet all work equally well, because the constraint is the single side, not the software.
The A3 report template has the same eight sections in the same order, with each section's question already written next to it.
The version with "11 jobs out of 24" tells you in four months whether anything has changed; the one with "delays have increased" does not. What remains is who writes those numbers and how long it takes.
How to fill in an A3 report: a single session, who sits at the table, which sequence
A sheet filled in by the business owner at their desk in twenty minutes, and a sheet filled in over an hour with the three people who live the problem: which of the two holds up when the verification date arrives?
The second, and not because of participation as such: the first contains the reconstruction of someone who was not there when the problem happened.
The A3 is filled in during a single session of forty-five to sixty minutes, with at least one person from each function involved and with the people who actually do the work sitting at the table.
The sequence is binding: section 5 does not open until section 4 has written down a cause, and section 3 is not written until section 2 has a number.
In Italy, running this session usually falls to the business owner, because among companies controlled by a person or a family, only 0.8% of units with 3 to 9 employees and 3.2% of those with 10 to 49 rely on an internal or external manager [5]: the person leading is also a party to the problem, and the way to neutralize the conflict is to ask questions instead of dictating answers [1].
This section covers who to invite, the order to follow and the five rules that keep the session within the hour.
- One side only, attachments kept separate.
- Fill it in order, from section 1 to section 8.
- The countermeasures section stays closed until section 4 has a written cause.
- Every number in section 2 carries its source: job log, delivery notes, tickets, recording sheets.
- Close by setting a verification date and an owner, not an intention.
When the business owner leads the session and is also a party to the problem, the move that keeps the table on the facts is to have the data written by the people who own it.
The data are gathered on site, through interviews with the people who run the process: this is the method documented in the Italian case study [3].
The sheet filled in over an hour with the people who live the problem holds up on the verification date because every number has an owner, and the quickest way to see this is to read one filled in from start to finish.
An example of a completed A3 report: drawing revisions after release to production
Before the sheet, the problem was called "engineering keeps changing the drawings mid-job"; afterward, there was a written cause and two countermeasures with a date: which of the two versions can be verified four months from now?
Only the second — and in the example the cause was not in the engineering department, but in the moment the job is released to production.
A hypothetical case built for teaching purposes: a company in an industrial engineering district that designs and builds made-to-order machines, thirty people across engineering, production, purchasing and sales, with the business owner personally following the largest jobs.
The problem is that drawing revisions arrive after the job has been released to production: eleven jobs out of twenty-four in the first half, with rework on the shop floor and supplementary purchase orders.
This section shows the sheet filled in section by section, separates the causes verified on the data from the hypotheses still open and arrives at two countermeasures with an owner and a verification date.
| Section | What was written | Why it is written this way |
|---|---|---|
| 1. Background | Jobs with variants have grown over the last year; engineering, production and purchasing are involved; the issue has come up in three meetings without a written decision | It pins down the scope and how long the problem has been going on |
| 2. Current situation | Revisions issued after release on 11 jobs out of 24 in the first half, 380 hours of rework on final accounts, 9 supplementary purchase orders, 6 delayed deliveries; source: job log and department final accounts | Every number has a period and a source that can be reopened |
| 3. Goal | Revisions after release on no more than 3 jobs out of 24 in the following half, measured on the same log | It is the hypothesis that section 7 will put to the test |
| 4. Cause analysis | Categories opened with the fishbone diagram, then a chain of whys on the heaviest branch; written cause: release to production opens on the sales order confirmation and not on the release of the bill of materials, and variants agreed verbally never make it back to engineering; verified on the data that 8 jobs out of 11 were released before the bill of materials, the seasonal peak remains a hypothesis | It separates what the data confirm from what is still to be verified |
| 5. Countermeasures | Three candidates weighed on impact, cost and reversibility with a decision matrix: (a) release to production opens only with a released bill of materials, (b) a variant form countersigned by the customer, (c) a fixed twenty-minute weekly check-in between engineering, purchasing and production; (a) and (c) selected, (b) stays in the queue because it affects the negotiation | The rejected options stay on paper too, with the reason |
| 6. Implementation plan | Release rule (engineering), weekly check-in (production), communication to sales (business owner), each with its own date | The names go here, not in section 4 |
| 7. Results verification | Same metric on the next 24 jobs, threshold of 3 or fewer, date four months out | Reviewing effectiveness is what clause 10.2 of the standard requires [7] |
| 8. Standardization | The release rule goes into the job management procedure, the weekly check-in into the calendar as a recurring event, the variant form stays under observation | What holds up becomes the normal way of working |
On the verification date, revisions are recounted on the new jobs: if the threshold is not met, you go back to section 4 with the data collected in the meantime, without having promised in advance how far the number would drop.
For a professional with outside collaborators, the three functions are client, collaborator and supplier, and the sequence does not change.
In a service company, the same structure applies to information that changes after the job has started.
The version that can be verified in four months is the second, the one with the written cause and two dates, and what derails it is a handful of mistakes that repeat from one company to the next.
The mistakes that reduce an A3 report to a form to be filed away (and how to avoid them)
A sheet filled in twenty minutes, and a sheet that stalls for an hour on section 4: which of the two produces a countermeasure that holds up?
The second, because the first has almost always skipped the step where the cause has to be written down — and a countermeasure without a cause is a decision made from memory.
The mistakes that hollow out an A3 look alike from one company to the next, and most of them consist of breaking the sequence.
The sheet is filled in by one person and brought to the others already finished; the current situation is written with adjectives instead of numbers; it is closed without a verification date, so the sheet becomes a document to be filed away.
This section lists them with the remedy alongside and with the signal that lets you recognize them while they are happening, before you have spent an hour in a meeting.
| Mistake | The signal that it is happening | The fix |
|---|---|---|
| Countermeasures open before the cause | Section 4 is shorter than section 5 | Read section 4 aloud before moving on |
| The sheet is filled in by one person | In the session, no one is there to correct the reconstruction | Invite the people who actually do the work |
| The current situation is written with adjectives | Section 7 does not know what to measure | A number, a period and the data source |
| People's names appear in the cause section | The tone of the session changes | Observable facts in section 4, names in section 6 |
| The sheet spills over into attachments | A second sheet is needed to contain it | The single side is the constraint; attachments stay separate |
| Verification has no date or no recorded outcome | There is no space to write the observed outcome [2] | Date and threshold before closing, plus an outcome line with its date |
The fourth mistake weighs more in a family business — 80.9% of Italian companies with at least 3 employees [5] — because it is what separates an analysis from a settling of scores.
The sheet that stalls for an hour on section 4 produces a countermeasure that holds up; the one filled in twenty minutes produces a decision made from memory: the mistakes are recognized while they happen, not once the sheet is filed.
Limits and conditions of applicability
The drawing revisions case is hypothetical and built for teaching purposes: the numbers serve to show how the sheet is filled in, not to estimate results.
The Italian case study [3] is a single case, a limit acknowledged by the authors: it describes a process carried out following the A3 model, not a measure of how well that model performs.
The A3 is not suited to an emergency in progress, where immediate containment comes before cause analysis.
Nor is it suited to a process governed by a single person, where the value of a document read in the same version by several functions disappears.
Without starting data in section 2, the sheet has nothing to verify in section 7, and verification becomes an impression.
The form has limited space for recording observed outcomes, a limit also documented in the literature [2]: the outcome line has to be planned before filling in the sheet, or kept on a supporting sheet.
Sources [1] and [7] were verified on their metadata and catalog record, not on the full text: the related statements are paraphrased, not quoted.
FAQ
What is an A3 report and why is it called that?
The A3 report is the format born at Toyota that keeps an entire problem on a single sheet, in the order it should be tackled: context, data, goal, cause, countermeasures, plan, verification, standardization [1].
The name comes from the 297 × 420 millimeter paper size, not from the method [2].
The short definition is in the glossary.
What is the difference between an A3 report and PDCA?
The A3 incorporates the logic of PDCA and adds what the cycle does not presuppose: a document that the functions involved read in the same version.
Is there an A3 template in Excel or Word?
A spreadsheet and a text document work as well as paper: what matters is that the eight sections stay readable all together, without references to attachments.
The A3 report template follows the same sequence, with the question each section answers written next to it.
What does a completed A3 report contain?
It contains the problem written with a number and a period, the cause verified on the data and kept separate from the hypotheses, the chosen countermeasures with an owner and a date, and the verification set for a future date.
The drawing revisions example above shows the eight sections filled in, from the starting numbers to the updated procedure.
How long does an A3 take?
The first pass usually takes between forty-five and sixty minutes, with the people involved at the table, plus a shorter session on the verification date.
This is an order of magnitude, not a promise.
Who should fill in the A3 report?
The people who live the problem fill it in, with a facilitator who asks questions instead of dictating answers [1].
The facilitator can be the business owner, as long as the data are written by the people who own them.
Operational summary
Before opening the sheet, run the problem through the entry filter: recurrence, measurable impact in time or money, risk of spreading, cost of doing nothing.
With two negative answers, the problem gets closed another, shorter way, and the sheet stays in the drawer.
The eight sections are filled in order, because each one needs what the previous one wrote: section 3 does not start without a number in section 2, and section 5 does not open without a cause in section 4.
Filling it in takes a single session, with at least one person per function involved and with the people who do the work at the table.
Every number in the current situation carries its source, and people's names go in the implementation plan, not in the cause analysis.
The sheet is closed by setting the verification date and the threshold before anyone gets up from the table, and on that date you recount the same numbers counted at the start.
What holds up at verification goes into the procedure; what does not hold goes back to cause analysis with the data collected in the meantime.
Conclusion
The A3 report is not a form to be filed away: it is the point at which a problem that cuts across several departments stops existing in three versions and starts existing in one, with a written cause and a verification date.
The sheet in the example works for a made-to-order machine builder just as it does for a practice with two collaborators: the names of the functions change, not the order.
When an A3 makes sense, and when another structured process is a better fit, is the subject of the guide to business problem solving.
The section where the analysis lingers longest has its own article, on root cause analysis.
A company that opens a sheet every time a problem comes back builds up an archive of its own recurring causes.
Meetings about "whose fault is it" give way to meetings about "what are we verifying," and the date promised to the customer depends on a rule, not on the memory of whoever was there.
Sources and references
[1] Sobek, D. K. II, Smalley, A., "Understanding A3 Thinking: A Critical Component of Toyota's PDCA Management System", Productivity Press, 2008 (ISBN 9781563273605; Shingo Research and Professional Publication Prize 2009). Available at: https://www.taylorfrancis.com/books/mono/10.4324/9781439814055/understanding-a3-thinking-durward-sobek-ii-art-smalley
[2] Bassuk, J. A., Washington, I. M., "The A3 Problem Solving Report: A 10-Step Scientific Method to Execute Performance Improvements in an Academic Research Vivarium", PLOS ONE, 8(10): e76833, 2013. Available at: https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0076833 — DOI: https://doi.org/10.1371/journal.pone.0076833
[3] Torri, M., Kundu, K., Frecassetti, S., Rossini, M., "Implementation of lean in IT SME company: an Italian case", International Journal of Lean Six Sigma, vol. 12, no. 5, 2021, pp. 944-972 (Politecnico di Milano — Department of Management, Economics and Industrial Engineering — for three of the four authors). Available at: https://www.emerald.com/ijlss/article-abstract/12/5/944/135997/Implementation-of-lean-in-IT-SME-company-an — DOI: https://doi.org/10.1108/IJLSS-05-2020-0067
[4] Baltrunaite, A., Formai, S., Linarello, A., Mocetti, S., "Proprietà, governance, management e performance delle imprese: evidenze dalle imprese italiane", Banca d'Italia, Questioni di Economia e Finanza no. 678, March 2022 (Invind survey, 2019 wave, about 3,200 manufacturing and service companies with at least 20 employees, MOPS section). Available at: https://www.bancaditalia.it/pubblicazioni/qef/2022-0678/index.html — PDF: https://www.bancaditalia.it/pubblicazioni/qef/2022-0678/QEF_678_22.pdf
[5] ISTAT, "Censimento permanente delle imprese 2023: primi risultati", press release, November 14, 2023 (reference year 2022; about 280,000 responding companies, representative of 1,021,618 companies with at least 3 employees). Available at: https://www.istat.it/comunicato-stampa/censimento-permanente-delle-imprese-2023-primi-risultati/ — PDF: https://www.istat.it/it/files/2023/11/REPORTCensimprese.pdf
[6] Accredia, "ISO Survey 2024: le imprese accelerano su sostenibilità e digitalizzazione", AccrediaHub, 2025 (data as of December 31, 2024). Available at: https://www.accredia.it/comunicazione/accrediahub/iso-survey-2024-le-imprese-accelerano-su-sostenibilita-e-digitalizzazione/
[7] UNI, "UNI EN ISO 9001:2015+A1:2024 — Sistemi di gestione per la qualità — Requisiti", Ente Italiano di Normazione, in force since October 16, 2024 (clauses 10.2 and 10.3). Available at: https://store.uni.com/uni-en-iso-9001-2015-a1-2024
[8] Treccani, entry "contromisura", Vocabolario on line, Istituto della Enciclopedia Italiana. Available at: https://www.treccani.it/vocabolario/contromisura/