Root-Cause Guide 9 min read

Root-cause analysis: the fishbone diagram and 5 Whys

A practical guide to the two tools that carry most of the root-cause load — the Ishikawa fishbone with its 6M categories for breadth, and the 5 Whys for depth — and how to combine them so a complaint is fixed at the cause, not the symptom.

Vidya Kathare · July 18, 2026 9 min read Updated July 2026
The 6M cause categories
1
Man
People, skill, training
Category
2
Machine
Equipment, tooling, wear
Category
3
Method
Process, procedure, setup
Category
4
Material
Inputs, components, supplier
Category
5
Measurement
Gauges, inspection, data
Category
6
Mother Nature
Environment, temp, humidity
Category

What root-cause analysis actually means

Root-cause analysis (RCA) is the disciplined search for the underlying reason a problem occurred — the cause that, if removed, stops the problem recurring — rather than the symptom you first see. A customer reports a leaking pump; the symptom is the leak, the immediate cause might be a failed seal, but the root cause could be a supplier who changed a rubber compound, a storage condition that hardened the seals, or an assembly torque spec that was never controlled. Treat the symptom and the pump gets a new seal and leaks again. Treat the root cause and the failure mode disappears.

RCA is the analytical heart of the investigate step in any complaint-handling lifecycle and of D4 in the 8D method. Two tools do most of the work and they are complementary: the fishbone diagram gives you breadth — it stops you fixating on the first plausible cause — and the 5 Whys gives you depth — it drives from a broad category down to the specific, actionable cause.

A symptom tells you what happened. A root cause tells you what to change. Confusing the two is how the same defect gets solved four times.

The fishbone (Ishikawa) diagram and the 6M

The fishbone diagram — also called the Ishikawa diagram after Kaoru Ishikawa, or the cause-and-effect diagram — organises the search for causes. The problem statement sits at the fish's head; a spine runs to it; and off the spine branch the major cause categories, each a "bone". The team brainstorms candidate causes onto the relevant bones, then tests the plausible ones against evidence. Its whole value is completeness: it forces the team to consider categories nobody wants to own, instead of anchoring on the first explanation offered.

In manufacturing the classic categories are the 6M:

Man

People and skill — training, competence, staffing, fatigue, handovers, unclear responsibility.

People

Machine

Equipment and tooling — wear, calibration, breakdown, incorrect setup, worn fixtures or dies.

Equipment

Method

Process and procedure — wrong or missing work instructions, uncontrolled parameters, bad sequence.

Process

Material

Inputs and components — out-of-spec raw material, wrong grade, supplier change, contamination.

Inputs

Measurement

Inspection and data — uncalibrated gauges, wrong method, sampling error, ambiguous acceptance criteria.

Inspection

Mother Nature

Environment — temperature, humidity, dust, vibration, lighting, storage and transport conditions.

Environment

The 6M is a manufacturing default, not a law. Service and transactional teams often use the 4S or 4P variants (Surroundings, Suppliers, Systems, Skills / People, Process, Policy, Plant). The categories are scaffolding to guarantee breadth — use whichever set fits your process, but pick one and use it consistently so your causes become analysable across cases.

How to build a fishbone in six steps

Building the diagram
1
Write a sharp problem statement
Put a specific, measurable effect at the head — "seal leak on Model X pumps, 4.2% of Q2 shipments", not "quality issues". A woolly head produces woolly bones.
2
Draw the spine and the category bones
Add one bone per cause category — the 6M, or your chosen set. This is the frame that keeps the team honest about breadth.
3
Brainstorm candidate causes onto each bone
Cross-functionally, ask "what under this heading could cause the effect?" Capture every plausible cause without debating yet — quantity first.
4
Drill each promising branch with 5 Whys
A first-level cause is rarely the root. Ask "why?" down each important twig until you reach something actionable and controllable.
5
Test the leading causes against evidence
Rank the candidates, then verify with data — check records, run trials, and where possible turn the problem on and off. A cause you cannot toggle is a hypothesis, not a root cause.
6
Confirm and hand off to corrective action
Lock the confirmed root cause (and escape point) and pass it to the corrective-action step. The diagram is a means to that decision, not the deliverable.

The 5 Whys — drilling from category to cause

The 5 Whys is deceptively simple: state the problem and ask "why?" repeatedly, each answer becoming the next question, until you reach a cause you can actually act on. "Five" is a rule of thumb — sometimes it is three, sometimes seven. Its discipline is refusing to stop at the first answer, which is almost always a symptom. The classic caution: 5 Whys can follow a single narrow chain and miss parallel causes, which is exactly why it works best inside a fishbone — the fishbone spreads the search wide, the 5 Whys drives the promising branches deep. And every "because" must be backed by evidence, or you have simply argued your way to a convenient conclusion.

A worked example, end to end

Complaint: a batch of machined housings shipped with burrs that fouled assembly at the customer. Fishbone brainstorming raises candidates across Machine (worn deburring tool), Method (no deburr check in the setup sheet) and Measurement (final inspection samples cosmetics, not burrs). The team drills the Method branch:

  • Why did burred parts ship? Because the deburr step was skipped on the night shift.
  • Why was it skipped? Because the deburr operation is not in the standard work instruction — it is "known" tribal knowledge.
  • Why is it not in the instruction? Because the instruction was never updated when the part was revised.
  • Why was it not updated? Because there is no trigger linking part revisions to work-instruction review.

The symptom was a burr; the immediate cause was a skipped step; the root cause is a missing change-control link between engineering revisions and shop-floor instructions. The corrective action is not "remind operators to deburr" — it is a controlled process that forces work-instruction review on every part revision, plus adding burr detection at inspection to close the escape point. That is a fix that spreads to every part, not just this one.

Fishbone vs 5 Whys — when to use each

AspectFishbone (Ishikawa)5 Whys
StrengthBreadth — many possible causes across categoriesDepth — drives one chain to the root
Best forComplex problems with several plausible contributorsSingle, well-scoped causal chains
Main riskStops at listing causes without proving anyFollows one path, misses parallel causes
Team sizeCross-functional group brainstormSmall group or individual, fast
Best usedTo open the investigation and structure itInside a fishbone branch, to go deep

They are partners, not rivals. Open with a fishbone to map the territory, then run 5 Whys down the branches that survive first contact with evidence — and always verify. See how this feeds the wider method in the 8D guide and becomes formal action in CAPA.

Root causes that live on the complaint, not in a spreadsheet

See a complaint escalate into a fishbone, cause categorisation and a proven-countermeasure lookup — all attached to the ticket that raised it.

Get a demo

How Fast Complaint Software supports root cause

Fast Complaint Software gives the investigate step real analytical depth instead of a free-text "resolution" box. A complaint captured as a numbered ticket escalates into the 8D and CAPA tooling, where a fishbone diagram and structured cause categorisation (cause category versus specification) organise the search, and a Problem Solving Report holds the narrative. A Past Trouble Database is checked first, so a new complaint is matched against proven countermeasures before the team reinvents one — many "new" problems are old problems with a different part number. Confirmed fixes spread through Horizontal Deployment, overdue investigations are chased by SLA follow-up and escalation, and the customer's post-resolution rating rolls into the Customer Satisfaction Index. The result is a root cause that is recorded, reusable and traceable to the complaint that raised it.

Frequently asked questions

What is a fishbone diagram used for?

A fishbone (Ishikawa or cause-and-effect) diagram is used to organise the search for the causes of a problem. The effect sits at the fish's head and the major cause categories branch off the spine as bones, so a team can brainstorm candidate causes across every category instead of fixating on the first explanation. It provides breadth in root-cause analysis.

What are the 6M categories in a fishbone?

The 6M are the classic manufacturing cause categories: Man (people and skill), Machine (equipment and tooling), Method (process and procedure), Material (inputs and components), Measurement (inspection and data), and Mother Nature (environment). Service teams often swap in variants like the 4S or 4P. The point is to guarantee breadth, so pick one set and use it consistently.

What is the difference between the fishbone and the 5 Whys?

The fishbone gives breadth — it lists many possible causes across several categories so nothing is overlooked. The 5 Whys gives depth — it drills a single causal chain down to the root by asking 'why?' repeatedly. They work best together: use the fishbone to map possible causes, then the 5 Whys to drive the promising branches to the actual root cause.

How many times should you ask why in a 5 Whys analysis?

'Five' is a guideline, not a rule. Keep asking why until you reach a cause that is genuinely actionable and controllable — sometimes that is three whys, sometimes seven. The discipline is refusing to stop at the first answer, which is usually a symptom, and backing every 'because' with evidence rather than assumption.

What is the difference between a symptom and a root cause?

A symptom is what you observe — the leak, the burr, the customer complaint. A root cause is the underlying reason it happened, the thing that if removed stops the problem recurring. Fixing the symptom (replace the seal) brings the problem back; fixing the root cause (control the compound and torque spec) removes the failure mode.

Is root-cause analysis required for ISO 9001?

Yes, in effect. ISO 9001 clause 10.2 requires you to determine the causes of a nonconformity and evaluate whether action is needed to eliminate those causes so it does not recur. Root-cause analysis with tools like the fishbone and 5 Whys is how organisations meet that requirement and produce evidence for an auditor.

Give every serious complaint a real root cause

Fast Complaint Software links a complaint to a fishbone, cause categorisation and a proven-countermeasure database — so investigations end in a confirmed cause, not a plausible guess.

Get a demo
No commitment. No slides. Your workflow on screen.