A3 Problem Solving: A Practical Guide to the 8 Steps (and Where AI Helps)
An A3 is one page that tells the whole story of a problem: what's wrong, why, what you'll do about it, and whether it worked. This guide covers what an A3 is, the 8 steps, how to write a strong problem statement, the mistakes that sink most A3s, and where AI can help without doing your thinking for you.
Practice in the free AI A3 Coach demo →
What is an A3?
An A3 is a structured problem-solving method that fits on one sheet of paper. The name comes from the paper size: A3 is the international size of 297 × 420 mm, close to an 11 × 17 inch tabloid sheet. The approach came out of Toyota and spread through Lean practice into nearly every industry.
The page matters less than the thinking. A good A3 makes you:
- state the problem in facts and numbers, not opinions
- find the real root cause before jumping to solutions
- tie every action to a cause, with an owner and a date
- check whether it worked, and lock in what you learned
Because it all fits on one page, an A3 is also a communication tool. A manager, a peer or a new team member can read it in a few minutes and see the logic, or spot the gap in it.
Why use an A3 instead of a slide deck?
Slide decks are good at looking finished. An A3 is good at showing whether the thinking holds up. One page forces you to choose what matters, and the fixed sequence of steps makes it hard to skip from "we have a problem" straight to "here's the fix."
That's the point. Most fixes that don't stick were aimed at a symptom, because nobody slowed down long enough to find the cause.
The 8 steps of A3 problem solving
Different organizations draw the boxes a little differently: some use seven, some nine. We use the 8-step practical problem-solving sequence, which is also how the A3 Coach app is built. Steps 1–5 are about understanding the problem, and steps 6–8 are about acting on it and making it stick. In PDCA terms (Plan–Do–Check–Act), steps 1–5 are Plan, step 6 is Do, step 7 is Check and step 8 is Act.
Step 1: Deviation from standard (problem identification)
The question: What should be happening, what is happening, and how big is the gap?
Name the target, name the current result, and name the gap, using a measure people already trust. Don't name a cause yet. If your problem statement already contains the answer ("we need more staff"), you've skipped the problem-solving.
Common mistake: a vague problem like "quality is bad" or "customers are unhappy." If you can't measure it, you can't tell whether you fixed it.
Step 2: Detailed problem description
The question: What does the data say when you break it down?
Quantify the problem and split it up. A Pareto chart (counts, percentages, cumulative percentage) usually shows that a few categories drive most of the gap. Drill into the biggest one. One A3 should solve one slice of the problem, not the whole thing.
Common mistake: trying to solve every type of defect, delay or complaint at once.
Step 3: Localization of origin
The question: Where in the process does the problem first appear?
Find the point of occurrence: the line, desk, queue, handoff or system step where things first go wrong. Go and look at the work itself. Don't just study a report about it. Teams often skip this step, and it's often the one that matters most.
Common mistake: doing root cause analysis on the whole process instead of the place where the problem actually starts.
Step 4: Containment and goal setting
The question: How do we protect the customer now, and what exactly are we aiming for?
Put in a short-term containment action so the problem stops reaching the customer or the next step while you work on the cause. Then set a clear goal, a target and a date, so everyone knows what "solved" means.
Common mistake: treating containment as the fix and stopping there.
Step 5: Root cause analysis
The question: Why does this happen, and can we prove it?
Brainstorm possible causes, often with a fishbone (Ishikawa) diagram. Prioritize them, then check the likely causes against facts. Only then ask "why?" again and again (5 Whys) to get from the direct cause to a root cause you can act on. Don't propose fixes in this step.
Common mistake: "root cause by opinion," where the cause is whatever the most senior person in the room believes.
Step 6: Countermeasures
The question: What will we do about each root cause, who owns it, and by when?
Write one countermeasure per root cause, each with an owner, a due date and a status. If an action doesn't connect to a confirmed cause, ask why it's on the page.
Common mistake: a long to-do list of good ideas that don't link back to the analysis.
Step 7: Observation and test results
The question: Did the countermeasures work?
Compare results with the goal from Step 4. Look for evidence, not hope. If the gap didn't close, go back to the analysis. That isn't failure. It's how PDCA works.
Common mistake: declaring victory as soon as actions are complete, before checking results.
Step 8: Standardize and follow-up (share the learning)
The question: How do we make the new way the normal way, and where else does this apply?
Update the standard (work instructions, checklists, system settings, training) so the gain holds. Then share what you learned with teams doing similar work.
Common mistake: solving the same problem again next year in another department.
How to write an A3 problem statement
A good problem statement is the hardest part of an A3, and the most valuable. Use this simple formula:
[Measure] was [actual] against a standard of [target], in [where], during [when].
Then check it: Is it measurable? Is it specific about where and when? Is it free of causes, solutions and blame?
| Weak (don't do this) | Stronger |
|---|---|
| "Customer support is too slow." | "In August, 38% of support tickets waited more than 24 hours for a first reply, against a standard of 10% or less." |
| "We need a new scheduling system." | "Over the last 6 weeks, 1 in 5 clinic appointments started more than 15 minutes late, against a target of 1 in 20." |
| "Line 2 has quality problems." | "Since March 1, Line 2 scrap is 6.1% of units against a standard of 2%; most of it is one defect type." |
Examples are made up to show the format. They aren't client results.
Notice that the stronger versions don't say why. That comes in Step 5.
The A3 layout: left side, right side
Most A3s read like a story, top to bottom and left to right:
- Left side: understand the problem. Background, current condition and the gap, problem breakdown, point of occurrence, goal, root cause analysis.
- Right side: act on it. Countermeasures with owners and dates, results against the goal, standardization and follow-up.
A rough rule: if the left side is thin, the right side is probably guessing.
A3 vs 8D vs PDCA vs DMAIC
These methods overlap a lot. They're all structured ways to solve problems. Here's when each one fits best:
| Method | What it is | Best for |
|---|---|---|
| PDCA | Plan–Do–Check–Act, the basic improvement cycle that the others build on | Quick, small problems and everyday improvement |
| A3 | PDCA told as a one-page story, usually owned by one person and coached by a manager | Internal problems that need clear thinking and alignment across a team or across functions |
| 8D | Eight Disciplines: a team-based method with formal containment and corrective action | Customer complaints and defects, especially where a formal corrective-action record is expected |
| DMAIC | Define–Measure–Analyze–Improve–Control, the Six Sigma project method | Data-heavy problems driven by process variation, usually run as longer projects |
You don't have to pick a side. Many teams use an A3 as the one-page summary of an 8D or DMAIC project.
The mistakes that sink most A3s
- Starting with the solution. The A3 turns into a justification for a decision that was already made.
- Scoping too big. One A3, one problem. Break big problems into slices.
- Skipping the "go and see." Writing an A3 at a desk from reports instead of looking at the work.
- Root cause by opinion. Causes nobody checked against facts.
- Countermeasures with no owner or date. If nobody owns an action, it won't get done.
- No follow-up. Steps 7 and 8 get skipped, so the problem comes back.
- Treating the template as the A3. Filling in boxes isn't problem solving. The thinking is the work.
Using AI for A3 problem solving
AI can make A3 work faster and more honest, as long as it coaches you instead of writing the A3 for you. Mistake #7 above applies here too: an AI that fills in every box just gets you to a finished-looking template sooner, with nobody having done the thinking.
Where AI helps
- Faster structure. AI can take you through the 8 steps in order, ask for what each step needs and keep the page organized, so the first draft takes less time to set up.
- Coaching questions on demand. AI can ask the questions a good coach asks, step by step, whenever you're working on the A3: "What's the standard, and what's actually happening?" "Where does the problem first appear?" "What fact confirms this cause?" "Which root cause does this countermeasure address?"
- Keeping the A3 honest. AI can point out the gaps that sink most A3s: a cause or a solution hiding in the problem statement, a root cause nobody checked against facts, a countermeasure with no owner or date.
- Clearer writing. AI can help tighten the wording so a manager can read the A3 in a few minutes. The facts still have to be yours.
Where people stay in charge
AI doesn't go and see the work, and it can sound confident and still be wrong. The A3 owner gathers the data, confirms causes against facts, picks the countermeasures and checks the results. Treat anything an AI suggests, especially a root cause, as a theory to check, the same way you'd treat a theory from anyone in the room. That's how we coach, too: We coach the thinking. You keep ownership of the work.
The A3 Coach app is built on that idea. It's an AI coach with a chat on the left and your live 8-step A3 on the right, and it won't let you skip ahead. Try it below.
Practice your first A3 with an AI coach, free
The fastest way to learn A3 problem solving is to do one on a real problem, with a coach who won't let you skip steps.
- Try the A3 Coach demo (free): an AI coaching chat on the left, your 8-step A3 on the right. Each step opens only once the one before it is built from facts. a3-coach-phase0.vercel.app
- Full A3 Coach app: $79/month. First month $39 with code FIRST39.
- One-to-one A3 coaching: work through a real problem with a coach. See A3 coaching or email hello@warneropex.ai.
Try the free AI A3 Coach demo → Talk to a coach →
Frequently asked questions
What is an A3 in Lean?
In Lean, an A3 is a one-page, structured way to solve a problem and share the thinking. It covers the problem, the current condition, the goal, the root cause, the countermeasures and the follow-up. It came out of Toyota and is used today in manufacturing, healthcare, services and tech.
Why is it called an A3?
It's named after the A3 paper size, 297 × 420 mm, about the same as an 11 × 17 inch tabloid sheet. The idea is that the whole problem-solving story should fit on one page.
Who invented A3 problem solving?
A3 problem solving came out of Toyota, which used A3-sized sheets to summarize problem solving, proposals and status reports. Lean practitioners and writers then spread it well beyond the auto industry.
What are the steps of A3 problem solving?
Templates vary from about seven to nine boxes. We use 8 steps: 1) deviation from standard (problem identification), 2) detailed problem description, 3) localization of origin, 4) containment and goal setting, 5) root cause analysis, 6) countermeasures, 7) observation and test results, 8) standardize and follow-up.
How do you write an A3 problem statement?
State the gap between standard and actual using a measure people trust, and say where and when it happens. For example: [measure] was [actual] against a standard of [target], in [where], during [when]. Leave causes and solutions out of it. Those come later in the A3.
How long does an A3 take?
It depends on the problem. A tightly scoped problem can move through the steps in days. A cross-functional problem often takes weeks, mostly because you're gathering data and waiting to see whether the countermeasures worked. If an A3 drags on for months, the scope is usually too big.
What is the difference between A3 and PDCA?
PDCA (Plan–Do–Check–Act) is the basic improvement cycle. An A3 is PDCA written as a one-page story: the early steps are Plan, countermeasures are Do, checking results is Check, and standardizing is Act.
What is the difference between A3 and 8D?
Both are structured problem-solving methods. 8D is team-based and is often used for customer complaints and defects where a formal corrective-action record is expected. An A3 is usually owned by one person, coached by a manager, and used for a wide range of internal problems.
Is A3 problem solving only for manufacturing?
No. A3 works anywhere there's a gap between what should happen and what does: service queues, clinics, offices, software teams. The steps stay the same. Only the data changes.
Can AI help with A3 problem solving?
Yes, if it coaches you rather than writing the A3 for you. AI can lay out the 8 steps, ask the questions a good coach would ask, and point out gaps such as a cause hidden in the problem statement or a countermeasure that doesn't link to a root cause. You still go and see the work, gather the facts and make the decisions. The A3 Coach app works this way: a coaching chat on the left, your live A3 on the right, and each step opens only once the one before it is built from facts.
Can AI do root cause analysis?
AI can help with root cause analysis. It can suggest possible causes, sort them into a fishbone diagram and keep asking "why?" It can't confirm a root cause for you. That takes data and a look at the actual work. Treat an AI-suggested cause as a theory to check against facts before you act on it.
Is there software for A3 problem solving?
Many teams use a spreadsheet or slide template. The A3 Coach app from Warner Consulting Group goes a step further: it's an AI coach that takes you through the 8 steps in a chat and builds the A3 as you go. The demo is free at a3-coach-phase0.vercel.app. The full app is $79 a month, with your first month $39 using code FIRST39.
Related
- Operational excellence consulting & coaching: how A3 fits into the bigger picture
- A3 coaching for leaders: one-to-one coaching on your real problem
- About Warner Consulting Group (Warner Opex)
Warner Consulting Group, LLC · Warner Opex · warneropex.ai