Insights
What an Ishikawa diagram is, how to set up the 6Ms, a template you can copy and a hypothetical worked example from a finance team.
By Christopher Fagan, Managing Partner, PEF Consulting Partners
Delivered for and alongside
















Fishbone diagram is the common name for a cause and effect diagram, a visual tool that sets out the possible causes of a problem under a handful of headings so a team can see and test them in one place. The problem sits at the head of the fish and the causes branch off the spine like bones.
It is one of the simplest tools in quality management and one of the most misused. Drawn well, with the right people in the room, it stops a group fixing on the first explanation offered. Drawn badly, it becomes a wall of sticky notes that confirms what everyone already thought.
In this article
The three names describe the same thing. Ishikawa diagram credits the man who popularised it, cause and effect diagram describes what it does, and fishbone describes what it looks like.
The anatomy is always the same. The effect, meaning the problem you are investigating, goes in a box at the right hand end of a horizontal line, the spine. Diagonal lines, the major bones, run off the spine for each category of cause. Individual causes hang off the major bones and sub causes hang off those, so the further you move from the spine, the more specific the cause becomes.
The diagram is named after Kaoru Ishikawa, the Japanese quality engineer who popularised it in the 1960s as part of the quality control movement in Japanese industry. He saw it as a tool for front line teams, particularly the quality circles he championed, rather than for specialists.
It became one of the so called seven basic tools of quality, alongside the check sheet, the histogram, the Pareto chart and the control chart, among others. Ishikawa’s Guide to Quality Control, published in English in 1976, set these tools out as a practical kit for working teams, and that is still the right way to think about the fishbone: a working tool, not a deliverable.
The classic manufacturing categories are the 6Ms:
Office, service and technology problems often fit better under headings such as people, process, systems, data, suppliers and policy. There is no prize for using a standard set. Choose four to six categories that suit the problem, and add one if the session keeps producing causes that have nowhere to go.
If the process itself is poorly understood, map it before you draw the fish. A SIPOC diagram of suppliers, inputs, process, outputs and customers takes an hour or so and makes the brainstorm far sharper.
The fish is easiest to draw on a whiteboard, but the record is easier to keep as a table. Copy the template below, write the effect at the top, and replace the typical causes with the ones your team raises. Mark each one as confirmed, ruled out or still to test.
Effect. State the effect precisely: what is going wrong, where, since when and by how much. Keep causes out of it.
| Category | Prompt questions | Typical candidate causes | Evidence and status |
|---|---|---|---|
| People | Do people have the skills, time, authority and information they need? | Training gaps, reliance on one person, unclear roles, conflicting priorities | Confirmed, ruled out or to test |
| Machine (systems and tools) | Do the systems and tools work as intended? | Manual workarounds, outages, slow batch jobs, poor integration | Confirmed, ruled out or to test |
| Method (process) | Is the process defined, followed and current? | Missing steps, handoffs, queues, rework, approvals that wait | Confirmed, ruled out or to test |
| Material (inputs and data) | Do inputs arrive on time, complete and correct? | Late or incomplete inputs, poor master data, supplier errors | Confirmed, ruled out or to test |
| Measurement | Are we measuring the right thing in the right way? | Inconsistent definitions, reports that double count, no baseline | Confirmed, ruled out or to test |
| Mother nature (environment) | Do timing, volume, location or regulation play a part? | Seasonal peaks, quarter end pressure, site layout, rule changes | Confirmed, ruled out or to test |
Keep the rejected causes on the record with the reason they were ruled out. If the problem returns, the next team starts from evidence rather than from scratch.
This example is hypothetical. The group, its figures and its systems are invented to show the tool in use.
A mid sized group with several trading companies takes nine working days to close its books each month, against a target of five. The finance director has been told the problem is the ERP system. Before committing to an expensive fix, she asks the finance team, the systems team and two budget holders to spend a morning on a fishbone, with the effect written as “Month end close takes nine working days against a target of five, and has done for six months”.
| Category | Candidate causes raised | What the evidence showed |
|---|---|---|
| People | One group accountant prepares all consolidation journals; budget holders slow to approve accruals | Confirmed: the group accountant’s tasks sit on the critical path for three days |
| Machine | Two ledgers since an acquisition; consolidation done in spreadsheets; slow overnight jobs | Confirmed: spreadsheet consolidation adds two days. Overnight jobs ruled out |
| Method | Intercompany balances reconciled only after period end; accruals rebuilt from scratch each month | Confirmed: intercompany mismatches take two to three days to clear |
| Material | Supplier invoices arrive late; purchase orders raised after the invoice | Partly confirmed: affects accruals but adds less than a day |
| Measurement | No tracking of close tasks; management pack reworked because KPI definitions differ between companies | Confirmed: rework adds a day after the ledgers close |
| Mother nature | Quarter end audit requests; holiday cover | Ruled out as a regular cause; affects some months only |
The session changes the conversation. The ERP system is part of the picture, because running two ledgers forces a manual consolidation, but most of the lost days come from how the work is organised: intercompany left until the end, one person on the critical path, and a management pack rebuilt every month because definitions differ. Those can be fixed in weeks, without a new system.
The team agrees three actions, each with an owner and a date: reconcile intercompany balances weekly during the month, split the consolidation journals between two people with a written procedure, and agree one set of KPI definitions for the pack. A month end close checklist gives them a task list to track progress against, and the ledger question moves into the longer term plan with a proper business case.
Notice what the fishbone did and did not do. It proved nothing on its own. It made sure every plausible cause was on the table, and gave the team a structure for checking each one against evidence before any money was spent.
Use one when a problem has several plausible causes and the people involved disagree, when you need a cross functional group to look at the same problem together, or when an investigation risks fixing on a favourite theory. It works well at the start of a root cause investigation, in a project review that needs to explain slippage, and in continuous improvement work.
It helps less when the cause is already obvious, or when the problem is a single technical fault that needs an engineer rather than a workshop. Where a failure involves complex combinations of events, fault tree analysis is a better fit. It is also a poor tool for choosing between options: it explains problems, it does not pick solutions.
A fishbone is the start of the work, not the end of it. The causes you confirm need owners, the risks it surfaces belong in a RAID log, and what you learn should go into a lessons learned template so the next team does not start from nothing.
On a struggling project the same tool works at a larger scale. In a project recovery, mapping why milestones are slipping under headings such as people, process, suppliers, technology, governance and data is usually quicker and more honest than reading the status reports. Where the same causes keep appearing across several teams, the answer is broader operational excellence work rather than another workshop, and in a finance function that often means looking at the wider finance transformation agenda.
PEF takes no reseller fees or commissions from any software vendor, so we have no reason to steer the technology bone towards a new system. Our Managing Partner, Christopher Fagan, has run systems and transformation delivery for 19 years, and our consultants test the causes with your team and stay to put the fixes in, rather than leaving a diagram on the wall.
A fishbone diagram, also called an Ishikawa or cause and effect diagram, is a visual tool for setting out the possible causes of a problem. The problem is written at the head, major categories of cause branch off a central spine, and individual causes hang off each category. It helps a team see every candidate cause before deciding which ones to test.
The 6Ms are the classic categories used in manufacturing: manpower (people), machine, method, material, measurement and mother nature, meaning the environment. They act as prompts so a team considers every kind of cause, not just the obvious one. Service and office teams often adapt them to headings such as people, process, systems, data, suppliers and policy.
Write a specific, measurable problem in the head of the fish. Draw a spine and add four to six category bones, such as the 6Ms. With the people who do the work, brainstorm causes under each category and ask why each one happens to add sub causes. Then mark which causes have evidence behind them and test the likeliest before acting.
The fishbone diagram is credited to Kaoru Ishikawa, the Japanese quality engineer, which is why it is also called an Ishikawa diagram. He popularised it in the 1960s through the Japanese quality control movement, and it became one of the seven basic tools of quality taught to front line teams and quality circles.
A fishbone diagram goes wide: it captures many possible causes across several categories at once. The 5 whys goes deep: it follows one chain of cause and effect down to a root cause. They work well together. Use the fishbone to surface candidates, data to narrow them, and then the 5 whys to drill into the few that matter.
Related reading: Root cause analysis: methods and template, SIPOC diagrams explained, Lessons learned template and Operational excellence consulting.
We run cause and effect sessions with your team, test the findings against data and stay to put the fixes in. Tell us about the problem and we will suggest a sensible first step.
Book a confidential conversation
Sources: Kaoru Ishikawa, Guide to Quality Control, 1976 (English edition); Nancy R. Tague, The Quality Toolbox, second edition, ASQ Quality Press, 2005
Written by Christopher Fagan
Managing Partner, PEF Consulting Partners
Christopher has led systems and transformation delivery for 19 years, including the National Grid separation, the £9bn Energy Bills Support Scheme and the BP Pulse EV charging rollout.
Where next