The Five Whys
A line has stopped, a team is waiting, and you hold the marker. Run the investigation yourself — and learn why the fifth answer fixes what the first one only patches.
Ask “why” five times.
The Five Whys comes from Taiichi Ohno, the architect of the Toyota Production System, and his canonical example starts with a machine that stopped. The obvious answer is right there on the panel: an overload blew the fuse. Most investigations end here — swap the fuse, restart, move on. Ohno’s discipline is to treat that first answer not as a conclusion but as the next question.
Each answer becomes the next question.
Why was there an overload? Because the bearing wasn’t sufficiently lubricated. Notice what just happened: the investigation dropped a level. The fuse was the symptom you could see; the dry bearing is the condition that produced it. Neither is the root — the chain keeps going.
Descend past the component to the system.
Why wasn’t it lubricated? Because the lubrication pump wasn’t pumping enough. Three levels down, and the question has moved from the machine that stopped to a completely different machine — the one whose quiet underperformance set the failure up. This is what asking why actually does: it walks you along the causal chain to places the symptom never mentioned.
Keep going while a “why” still has an answer.
Why wasn’t the pump pumping? Its shaft was worn and rattling. A worn shaft could still tempt you to stop — “replace the shaft” sounds like a fix. But a replaced shaft wears out again unless you know why it wore. One more level.
The root: a missing strainer.
Why was the shaft worn? There was no strainer on the pump, so metal scrap got into the lubrication oil and ground the shaft down. Five whys from a blown fuse to a missing filter — a cause nobody would have guessed from the fuse panel, and one you can verify by looking.
Two fixes, categorically different.
Compare the countermeasure at the top of the chain with the one at the bottom. Replace the fuse and the machine stops again — the scrap is still in the oil, the shaft is still wearing. Fit a strainer and the entire chain above it becomes impossible. That is the whole point of the method: the depth of the question decides the category of the fix.
Now you hold the marker.
Tuesday, 10:40 a.m. Packaging Line 2 has stopped, a shift lead is standing at the case sealer, and the team has gathered a whiteboard. You hold the marker. At each level, three answers are on offer — only one of them is a verified fact you could go see. Write that one down.
The Ohno chain
The machine stopped.
- Why did the machine stop? There was an overload and the fuse blew.
- Why was there an overload? The bearing was not sufficiently lubricated.
- Why was it not lubricated sufficiently? The lubrication pump was not pumping sufficiently.
- Why was it not pumping sufficiently? The shaft of the pump was worn and rattling.
- Why was the shaft worn out? There was no strainer attached and metal scrap got in.
Replace the fuse: The machine stops again in a few months — same fuse, same reason. Attach a strainer to the pump: Scrap never reaches the shaft. The whole chain above it never happens again.
The investigation
Tuesday, 10:40 a.m. Packaging Line 2 has stopped, a shift lead is standing at the case sealer, and the team has gathered a whiteboard. You hold the marker. At each level, three answers are on offer — only one of them is a verified fact you could go see. Write that one down.
Why #1: Why did the line stop?
- The operator wasn’t paying attention when the carton went in.
That’s a verdict about a person, not a fact about the process. Ask why of the process, not the person — the operator was doing the same job the same way as every other shift. If you write a name on the board, the only next question is “why is this person bad?”, and that question has no useful answer. Look at what physically happened instead.
- We need a backup sealer so a jam can’t stop the whole line.
That’s a countermeasure, not a cause — and it arrived before anyone knows what the problem is. A second sealer would jam on the same cartons for the same reason. Hold the solution; first write down what you can actually observe about why the line stopped.
- The case sealer jammed on a carton.
Confirmed at the machine: a carton is wedged in the flap folder, and the log shows this is the fourth jam this month. A fact you can point at — write it down and keep digging.
Go see: You walk to the sealer. A half-crushed carton is wedged in the flap folder — the jam is right there to see, and the machine log shows three more like it this month.
Fishbone category: Machine — Sealer jammed
Why #2: Why did the sealer jam?
- The carton flaps were warped and wouldn’t fold flat.
Confirmed in your hands: the jammed carton and several behind it have bowed flaps. Flat cartons run fine; warped ones catch in the folder. The machine isn’t the problem — what’s being fed into it is.
- Maintenance hasn’t been keeping the sealer tuned properly.
You’ve moved the blame one desk over, but it’s still blame — and it isn’t what you saw. The sealer folds flat cartons all day without complaint; the maintenance log is clean. Before pointing at a team, check what the machine was actually being asked to seal.
- Our equipment is getting old.
Too abstract to act on. “Old equipment” explains everything and therefore nothing — you can’t go see “old,” and the fix it implies is a capital request, not an investigation. Stay concrete: what, physically, was the sealer doing when it jammed?
Go see: You pull the jammed carton out and check the ones queued behind it. The flaps on several are visibly bowed — they won’t sit flat, so the folder catches them mid-stroke.
Fishbone category: Material — Warped cartons
Why #3: Why were the cartons warped?
- We should switch to a better carton supplier.
A solution again, and aimed at the wrong target. The rest of the delivery — same supplier, same batch — is flat and runs fine. Before you fire a supplier, follow the warped cartons backwards: where exactly did this pallet come from, and where has it been sitting?
- That pallet sat by the open dock door and absorbed moisture.
Confirmed at the dock: the pallet the warped cartons came from was stored beside the open door through a week of rain. Cartons from pallets stored inside are flat. Same supplier, same batch — different storage spot.
- Suppliers just aren’t as reliable as they used to be.
That’s a mood, not a cause. It names no supplier, no pallet, no mechanism — nothing you could verify or fix. The cartons in front of you got warped somewhere specific. Go find the somewhere.
Go see: You trace the warped cartons back to their pallet. It was sitting beside the open loading-dock door — after a week of rain, the outer layers of corrugate have taken on moisture and bowed.
Fishbone category: Environment — Dock-door humidity
Why #4: Why was the pallet stored by the dock door?
- The receiving team was careless about where they put it.
Go watch a delivery arrive before you write that. Staging was full; the team put the pallet in the only open space they had. Anyone on any shift would have done the same, because nothing tells them where overflow should go. When every reasonable person would make the same “mistake,” the cause is in the process, not the people.
- We need to train receiving on proper storage practices.
Training on what, exactly? There is no overflow standard to train anyone on — that’s the finding. “More training” is the most common way to close an investigation early: it sounds like action and changes nothing. Keep asking why the staging area couldn’t hold the delivery in the first place.
- The staging area was full — there was no marked place for overflow.
Confirmed with the receiving team: staging was full when the truck arrived, and no standard says where overflow goes. The pallet landed in the only open space — which happens to be the worst spot in the building for corrugate.
Go see: You ask the receiving team to walk you through the last delivery. The marked staging area was already full, so the overflow went to the only open floor space — beside the dock door. There is no rule for where overflow goes; there has never needed to be one.
Fishbone category: Method — No overflow standard
Why #5: Why was the staging area full?
- Management doesn’t invest in enough warehouse space.
You’ve floated up to the org-chart level, where every problem becomes “management” and every fix becomes someone else’s job. Maybe the building is too small — but you haven’t checked why this area overflows. Look at what fills it, and on what rhythm, before concluding the walls are in the wrong place.
- The planner keeps over-ordering cartons.
The planner is following the ordering policy, not defying it — the monthly bulk buy is the company’s own rule, written to capture a volume discount. Blaming the person executing a policy protects the policy. Ask what the policy asks of the space instead.
- Cartons are ordered monthly in bulk — a month of stock, a week of space.
Confirmed in the records: one bulk order a month, sized for the discount, arriving into a staging area sized for a week. The overflow isn’t an accident — it’s scheduled. You’ve hit a decision, not an event. This is the root.
Go see: You pull the ordering records with the planner. Cartons are bought once a month to hit a volume discount — a month of stock arrives at once, into a staging area sized for about a week. It has overflowed on delivery week every month this year.
Fishbone category: Method — Monthly bulk ordering
The countermeasure ladder
- Restart the line. It runs — until the next warped carton. You’ll be back this afternoon.
- Clear the jam and add a “watch the sealer” reminder. The fourth jam this month becomes the fifth. Nothing upstream changed.
- Inspect cartons and scrap the warped ones before the line. The jams slow down — and you’ve added an inspection job and a scrap bin, forever, to absorb a problem that keeps arriving.
- Move that pallet away from the dock door. Next month’s delivery overflows into the same spot, because the space problem that put it there is untouched.
- Mark an overflow location away from the door, with a FIFO rule. Genuinely better — the moisture path is closed. But staging still overflows every delivery week, and overflow always finds a way to hurt you.
- Change the ordering rhythm: smaller, more frequent carton orders sized to the space (and re-price the discount against a month of jams, scrap, and stops). Staging never overflows. Nothing gets stored by the door. Cartons stay flat. The sealer stops jamming. The chain above this line never happens again.
The three ways an investigation goes sideways
- Blame ends investigations. The person in front of a problem is your best witness, not your suspect — when every reasonable person would have done the same thing, the cause lives in the process.
- A solution proposed before the cause is known is a guess wearing a hard hat. Hold every countermeasure until you can say what it counters.
- “Management,” “culture,” and “old equipment” explain everything and fix nothing. A usable cause is specific enough to go see — a pallet, a door, an ordering rule.
What the chain can’t do alone
When one chain misleads
The Five Whys follows a single thread, and real problems are sometimes a braid. If the line stops when cartons are warped and the sealer’s pressure drifts and a new operator rotation started the same week, one chain will find one of those and quietly present it as the whole story. The tell is a chain that feels true but doesn’t fully explain the pattern — the problem recurs on days your root cause was absent. When you suspect multiple contributing causes, map the width before you dig the depth: put every candidate on a fishbone diagram — Method, Machine, Material, People, Measurement, Environment — and then run a why-chain down each bone that carries real evidence. The two tools aren’t rivals; the fishbone is the map, the Five Whys is the drill.
Every “why” is a claim — go verify it
The most dangerous place to run a Five Whys is a conference room, because every level will get an answer whether or not anyone knows it’s true. Five confident guesses in a row compound into a root cause that’s pure fiction — with a countermeasure to match. In the investigation you just ran, every confirmed answer came with a “go see”: the jam you could touch, the bowed flaps, the pallet by the door, the ordering records. That’s the standard. Treat each answer as a hypothesis until someone has stood where the work happens and checked — the same go-and-see discipline you can practice in the gemba walk guide. A why-chain is only as strong as its weakest unverified link.
From root cause to action: the A3
A confirmed root cause is the middle of the story, not the end. The ordering-rhythm decision you uncovered still needs an owner, a target condition, countermeasures someone agrees to, and a follow-up date — and the people who set that ordering policy need to see the chain of evidence, not just your conclusion. That’s what an A3 is for: the Five Whys you just ran is precisely the analysis box of an A3 problem-solving sheet, with the background and current condition above it and the countermeasures and follow-up below. Drill with the whys; carry the result on an A3.
Packaging Line 2, its shift lead, and the carton chain are an honest hypothetical, created for teaching — no real company or plant is depicted, and no figures in this guide are real statistics. The machine-overload example follows Taiichi Ohno, Toyota Production System: Beyond Large-Scale Production (Productivity Press, 1988).
Take the marker at your next line stop.
Train your team on this
AI-built lean training that puts your team through investigations like this one — for your industry, your processes.
kaizumi.com/lean-training/tryMap all the bones
The free fishbone diagram builder — put the width of the problem on one page before you drill.
kaizumi.com/tools/fishboneFrequently asked
- What is the Five Whys technique?
- A root cause analysis method from Taiichi Ohno and the Toyota Production System: when a problem appears, ask "why did this happen?" — then ask "why?" of that answer, and keep going until you reach a cause you can fix so the problem cannot recur. Ohno's classic example descends from a blown fuse (overload) to insufficient lubrication, to a pump not pumping, to a worn pump shaft, to the real root: no strainer on the pump, so metal scrap got into the oil. The fix at the bottom — fit a strainer — prevents everything above it; the fix at the top — replace the fuse — only resets the clock.
- Is it always exactly five whys?
- No — five is a rule of thumb, not a rule. Some chains reach a fixable root cause in three questions; others need seven. The real stopping test is: have you reached a cause that, if countered, prevents the problem from recurring — and is it something you control? Stop too early and you get a symptom patch (replace the fuse); push past the useful root and you drift into abstractions like "human nature" or "the economy" that no one can act on.
- What is the difference between Five Whys and a fishbone diagram?
- Depth versus width. The Five Whys drills one causal chain deep — each answer becomes the next question, down a single path. A fishbone (Ishikawa) diagram maps the width first: every candidate cause laid out across category bones (Method, Machine, Material, People, Measurement, Environment). They work best together: when a problem may have multiple contributing causes, map the width with a fishbone, then run a why-chain down each bone that carries real evidence. One five-whys chain typically touches only two or three of the six bones — the rest stay unexplored unless you go map them.
- How do you keep a Five Whys from blaming people?
- Ask why of the process, not the person. The moment an answer names a person — "the operator wasn't paying attention" — the investigation dies, because the only next question is "why is this person bad?", which has no useful answer. The discipline: if every reasonable person in the same situation would have done the same thing, the cause is in the process. Rewrite the answer as what physically happened, and keep asking why the process allowed it.
- Why do Five Whys investigations fail?
- Three classic failure modes: blaming a person instead of examining the process; jumping to a solution before the cause is known ("we need more training" is the most common way to end an investigation early); and going abstract ("management doesn't care") — an answer too vague to verify or fix. A fourth killer is running the whole exercise in a conference room: each unverified answer is a guess, and five guesses compound into a fictional root cause. Verify every answer at the gemba — go see the jam, the pallet, the records — before writing it down.
- What do you do after finding the root cause?
- A confirmed root cause is the middle of the story, not the end. It needs a countermeasure with an owner, a target condition, and a follow-up date — and the people who own the process you're changing need to see the chain of evidence, not just the conclusion. In practice the Five Whys is the analysis box of an A3 problem-solving sheet: background and current condition above it, countermeasures and follow-up below.
Founder of Kaizumi, an AI-powered Lean training platform. More about Matthew →
Updated July 2026 · Packaging Line 2, its shift lead, and the carton chain are illustrative, created for teaching — an honest hypothetical, not a case study. The machine-overload example follows Taiichi Ohno, Toyota Production System: Beyond Large-Scale Production (Productivity Press, 1988).