Skip to content
LeanSuite

Lean Studio board · Problem solving

Fishbone diagram template (Ishikawa, 6M)

In one sentence

A fishbone diagram, also called an Ishikawa or cause-and-effect diagram, sorts every possible cause of one problem into categories, usually the six M's, so a team sees where the causes cluster before it tests any of them. A fishbone diagram template gives you that chart ready drawn, with the bones labelled.

This free fishbone diagram template comes three ways: a printable PDF for a session at the line, an Excel cause list, and a Lean Studio board for a team on screens. The Ishikawa diagram template has the problem statement at the head and six bones for the six M's: man, machine, method, material, measurement and mother nature. Page 2 of the PDF is the same sheet filled in for a scrap problem on a stamping line, so you can see what good looks like before you start. Kaoru Ishikawa, who first used the diagram in 1943, describes it in his Guide to Quality Control.

Free to use: print it, copy it and edit it for your team. Enter your name and work email once to download.

The problem 'scrap on Line 2 up to 4.1%' at the head and stickies with candidate causes on the six M bones, such as a worn die and an uncalibrated gauge. Sample data.

When to use it

When to use the Fishbone (Ishikawa) board

  • At the start of a root cause analysis, once the problem is defined and before 5 Whys narrows one cause.
  • When a problem has many possible causes and the team keeps arguing about which one matters.
  • In a kaizen event or an 8D team meeting, to get every idea on the wall in ten minutes.
  • For recurring defects, downtime or safety events where the first fix did not hold.

What is on the board

  1. 1

    Problem statement (the head)

    One measured problem, for example scrap on Line 2 up from 1.8% to 4.1% since March.

  2. 2

    Man

    Training, staffing, handover, fatigue.

  3. 3

    Machine

    Equipment, tooling, settings, wear.

  4. 4

    Method

    The procedure, the standard, the sequence.

  5. 5

    Material

    Raw material, components, supplier lots, storage.

  6. 6

    Measurement

    Gauges, inspection method, how the data is collected.

  7. 7

    Mother nature

    Temperature, humidity, lighting, layout.

The fishbone diagram templateA problem statement box at the head, six bones labeled man, machine, method, material, measurement and mother nature with blank lines for causes, a how-to-use strip and a table for the causes to check.
Problem statement headSix M bonesVotes and top causes to checkConfirmed or ruled out, with evidenceEmpty-bone checkFilled-in sampleHow to verify a causeExcel cause list
Filled-in Fishbone diagram template (6M): scrap on a stamping line, 15 causes and 4 checks
The same sheet filled in with the worked example (page 2 of the PDF, illustrative numbers). Open full size (opens the picture in a new tab)

Free template

Free fishbone diagram template: printable PDF and Excel cause list

The same six-M fishbone on a landscape sheet for a session on paper, with a table for the causes you will check. Page 2 is this board's sample filled in: 15 causes, 21 votes, the top 3 checked against the scrap log and one confirmed. Page 3 covers how to pick and check causes, five traps and the source. The Excel cause list ranks the votes, picks the top causes to check, flags empty bones and causes marked confirmed with no evidence, and totals each bone.

Free to use: print it, copy it and edit it for your team. Enter your name and work email once to download.

How to fill in the template

  1. 1
    Write the problem in the head as one measured sentence: what, where, how much and since when. The sample reads: scrap on line 2, up from 1.8% to 4.1% of parts made, since early March.
  2. 2
    Fill the bones in a short, silent round, one cause per line, then read each bone aloud and move any cause that sits on the wrong bone. Rename a bone if your process needs another heading.
  3. 3
    Give everyone three votes and mark them as dots beside the causes they think most likely. Seven people make 21 dots.
  4. 4
    Copy the causes and votes into the Excel cause list. It ranks them and marks the top three, or the number you set, as the causes to check.
  5. 5
    Give each cause to check a method, an owner and a date. The sheet flags a check that is past its date.
  6. 6
    Write the result as confirmed or ruled out, with the evidence. The sheet flags a cause marked confirmed with no evidence and any bone left empty.

Running it

How to run a Fishbone (Ishikawa) session

  1. 1

    Write the problem as one measured sentence and put it in the head before anyone names a cause. If the team cannot agree on the sentence, it is not ready for a fishbone.

  2. 2

    Bring the people who touch the process: an operator from each shift, the setter, maintenance, quality. Five to eight people, typically an hour; a manager in the room should ask questions, not fill bones.

  3. 3

    Brainstorm in silence for five minutes, one cause per sticky, then place the stickies on the bones. Ask 'why does that happen?' of any sticky that is really a symptom.

  4. 4

    Read each bone aloud and move the stickies the team disagrees with. Merge duplicates rather than deleting them: two people naming the same cause is information.

  5. 5

    Pick the two or three causes to test, by data if you have it, by dot vote if you do not, and mark them with a different sticky colour.

  6. 6

    Give each chosen cause a person and a date to check it on the floor: measure it, watch it, pull the record. The fishbone ends when the causes to test have names against them.

  7. 7

    Good output: 15 to 30 causes, every bone with something on it, three causes marked, three owners, and a date for the 5 Whys or the test.

Worked example · sample data

Worked example

The sample board has 'Scrap on Line 2 up to 4.1%' at the head. Under Man the team put a new operator on nights and a skipped handover; under Machine, a worn die on press 2 and clamp pressure that drifts; under Method, two versions of the setup sheet; under Material, coil lot 4471 running thinner than spec; under Measurement, a gauge past its calibration date; under Mother nature, a shop floor at 34°C in July. Seven people voted and checked the top three against the scrap log. Lot 4471 ran at 8.6% scrap against about 2% on the other lots: confirmed. The gauge was within tolerance and both setup sheets held the same settings: ruled out. The heat was ruled out by timing, since the rise began in March. The team booked a 5 Whys on how a thin lot passed incoming inspection.

The board shown at the top of this page, read through. Invented data, not a customer.

15 causes, 21 votes, bone by bone

  1. Man4 votes

    • New operator on nights2 votes
    • Handover skipped between shifts2 votes
    • Setter covering two linesno votes
  2. Machine3 votes

    • Worn die on press 22 votes
    • Clamp pressure drifts1 vote
    • Die lubrication nozzle partly blockedno votes
    • Press 2 feed rollers wornno votes
  3. Method5 votes

    • Two versions of the setup sheet in circulation4 votesRuled outBoth versions hold the same settings for the 3 parts run since March
    • No first-off check after a coil change1 vote
  4. Material6 votes

    • Coil lot 4471 thinner than spec5 votesConfirmed8.6% scrap on lot 4471 against 1.7% to 2.1% on the other lots; 10 of 10 samples below the thickness limit
    • New lubricant supplier since February1 vote
  5. Measurement3 votes

    • Gauge past its calibration date3 votesRuled outWithin tolerance as found; a second gauge gave the same scrap count
    • Scrap now counted at the press, not at final inspectionno votes
  6. Mother natureno votes

    • Shop floor at 34°C in Julyno votesRuled outThe rise started in March, months before the heat
    • Coils stored next to the open dock doorno votes

one voteStruck through: checked and ruled out.

Man, 4 votes: New operator on nights (2 votes); Handover skipped between shifts (2 votes); Setter covering two lines (no votes). Machine, 3 votes: Worn die on press 2 (2 votes); Clamp pressure drifts (1 vote); Die lubrication nozzle partly blocked (no votes); Press 2 feed rollers worn (no votes). Method, 5 votes: Two versions of the setup sheet in circulation (4 votes, ruled out); No first-off check after a coil change (1 vote). Material, 6 votes: Coil lot 4471 thinner than spec (5 votes, confirmed); New lubricant supplier since February (1 vote). Measurement, 3 votes: Gauge past its calibration date (3 votes, ruled out); Scrap now counted at the press, not at final inspection (no votes). Mother nature, no votes: Shop floor at 34°C in July (no votes, ruled out); Coils stored next to the open dock door (no votes). The top 3 by votes were Coil lot 4471 thinner than spec, Two versions of the setup sheet in circulation, Gauge past its calibration date.

7 people with 3 votes each make 21 votes, and the 15 causes above carry 21. The top 3 by votes were checked against data; the heat was ruled out by timing. Only coil lot 4471 thinner than spec was confirmed, so the votes set the order of the checks and the scrap log found the cause. Page 2 of the PDF and the Excel example sheet hold the same rows.

Common mistakes

  • A vague head. 'Quality problems on Line 2' produces a fishbone nobody can test; a number and a date do.
  • Filling the bones from the meeting room. Walk the line first, or bring the people who work it.
  • Treating the fishbone as the analysis. It lists candidates; the root cause still has to be proven on the floor.
  • Letting one loud voice decide which cause is right before the board is full.
  • Leaving a bone empty and moving on. An empty bone usually means nobody in the room knows that part of the process.
  • Never coming back. Update the board when a tested cause is ruled out, so the team does not test it twice.

How it works in Lean Studio

  1. 1
    Start from the Fishbone (Ishikawa) template and type the problem into the head.
  2. 2
    Add a sticky per candidate cause and drop it on its bone; drag it to another bone when the team changes its mind.
  3. 3
    Mark the causes to test next with a different sticky colour, then take them to a 5 Whys board.
  4. 4
    Save the finished board to My templates so the next team starts from your version.

Lean Studio boards are drawn with stickies, text, shapes (rectangle, rounded rectangle, ellipse, diamond), arrows and charts, and any board can carry the 15 data widgets, from a scorecard to a kanban. See all widgets and boards.

Also in LeanSuite

Where it goes next in LeanSuite

LeanSuite's RCA Agent drafts 5 Whys, 4M1D and fishbone analyses from a problem and its attached documents.

  • The causes to test become actions with owners and due dates in a Kaizen project, which has a fishbone section of its own alongside 5 Whys, 5W1H and 4M1D. Kaizen & Projects

  • If the problem was raised as a tag on the floor, a recurring tag can be escalated into a Kaizen or A3 project in one click, with its history. Daily Issue Management

  • The RCA Agent drafts a fishbone from the problem description and attached SOPs or manuals, for the team to edit. AI Root Cause Analysis

See AI Root Cause Analysis

FAQ

Fishbone (Ishikawa) board: common questions

  1. This one comes as a PDF to print and an Excel file. The Excel file is a cause list rather than a drawing: it ranks the votes, marks the causes to check and tracks each one to confirmed or ruled out. For a drawn diagram to paste into a Word report or a slide, type the causes into the free fishbone diagram maker and download it as a PNG or SVG.

  2. Man (people), machine, method, material, measurement and mother nature (environment). Some plants add management and maintenance to make eight. Office and service teams often use the four P's instead: policies, procedures, people and plant. Pick categories that cover the whole process and keep to six or so, or the diagram turns into a list.

  3. Fishbone first. It spreads the possible causes wide, then 5 Whys drills one of them down to a root cause. Doing 5 Whys first tends to chase the first cause somebody names.

  4. Use it to list the suspects, then prove one. Vote or use data to pick the two or three causes to check, then split the defect data by each one (lot, shift, machine), check the cause started when the problem did, and where you can, put it back and take it away. Mark each cause confirmed or ruled out and write the evidence. Run 5 Whys on the confirmed cause.

  5. As many as the team can name in a short, timed session, usually 15 to 30. Then vote or use data to pick the two or three worth testing. A bone with nothing on it is a prompt to think harder, not a finding.

  6. Yes. Lean Studio is a shared whiteboard, so people add stickies from their own screens and the board updates for everyone. Teams without LeanSuite can share the Excel cause list instead.