White Belt · 12 min
5S: why the tidy area keeps going back
After this you can
- conduct a 5S audit of a workspace and score it against each of the five stages.
Assumes you have done Go and see: what a walk gives you that a report cannot.
The problem
A goods-in bay had been running a 5S programme for eight months and scored 4 out of 4 for Sustain — the audit happened every month, on time, with a signed sheet. It scored 1 for Sort: the same obsolete pallet jacks had been on the racking at all eight audits. The score was measuring the reliability of the audit routine. Nobody had noticed, because the total looked respectable and the total was the only number anyone read.
The idea
5S is five stages, in order, and the order is not decorative — each one depends on the ones before it.
Sort (seiri) — remove what is not needed for the work done here. The test is not what looks untidy; it is what you would have to move to do the job.
Set in order (seiton) — everything needed has a marked place, and a person new to the area can find it without asking. The newcomer is the test, not the person who has worked here for fifteen years and knows where everything is.
Shine (seiso) — cleaning as inspection. This is the stage most often reduced to tidying, and the reduction removes the whole point: cleaning a machine is how you find the leak, the wear, the missing fastener. If cleaning is not revealing problems, it is housekeeping.
Standardise (seiketsu) — make the first three the normal condition, with the standard visible at the point of use. A folder in the supervisor's office is not a standard anybody works to.
Sustain (shitsuke) — the condition holds without an audit being announced. This is the one that fails, universally, and it fails because it is treated as a matter of discipline rather than of design.
The failure the audit score hides
Sustain cannot be in better shape than the things it sustains. It holds the first four in place; a high Sustain score over weak foundations is a description of an audit routine, not of a workplace. The same is true one stage earlier: Standardise documents a condition, and documenting one does not create it. A pristine standard for a layout that no longer exists scores well and means nothing.
So when you read a 5S score, read the shape before the total. A score of 1-1-2-3-4 rising through the stages is upside down: it says the paperwork is in better condition than the workplace.
Score with an observation, or do not score
A number on its own is an impression. It cannot be compared with next month's, and it tells the area nothing about what to change. "Sort: 1 — obsolete pallet jacks still on the racking, as at the last three audits" is an audit. "Sort: 1" is a mood.
The observation is also what makes the score reproducible. If a different auditor could read your sentence and arrive at the same number, the scale means something. If not, you have a column of numbers that only compare with each other by accident.
And never compare areas
A 5S total compares an area with itself over time. It does not compare two areas, because the scale is anchored by what a particular auditor has seen, and two auditors in two buildings are not using the same one. Ranking areas by 5S score reliably produces the one outcome 5S cannot survive: people scoring for the league table rather than for the workplace.
Worked example
The goods-in bay, audited:
| Stage | Score | What was observed |
|---|---|---|
| Sort | 1 | Obsolete pallet jacks still on the racking, as at the last three audits |
| Set in order | 1 | (nothing recorded) |
| Shine | 2 | Floor clean; hydraulic seep on jack 4 not recorded anywhere |
| Standardise | 3 | Standard on the wall, dated, and shows the pre-March layout |
| Sustain | 4 | (nothing recorded) |
Total: 11 out of 20. Middling. Unremarkable. It is also the least informative number on the page, and it is the one that gets reported.
Read the shape instead. The scores rise through the stages — 1, 1, 2, 3, 4 — which is exactly backwards. Every stage depends on the ones before it, so a healthy area looks like 4, 4, 3, 3, 2: strong foundations, and the difficulty concentrated at the far end where it belongs.
Sustain 4 against Sort 1. Sustain keeps the earlier stages in place. It cannot be at 4 while what it is holding is at 1 — unless what is being sustained is the audit rather than the workplace, which is precisely what has happened here. Eight audits, all on time, all recording the same pallet jacks.
Standardise 3 against Sort 1. The standard is on the wall, dated, and describes a layout that changed in March. It scores well because a standard exists. Nobody asked whether it matches the area, and nobody could have worked to it if they had tried.
Two scores have no observation. Set in order at 1 and Sustain at 4 are both unsupported. The Sustain 4 in particular cannot be checked, cannot be reproduced by another auditor, and cannot be compared with next month's — which is unfortunate, because it is carrying more than a third of the total.
What Shine actually found. Score 2, and the sentence is the most valuable line on the sheet: the hydraulic seep on jack 4 is not recorded anywhere. Cleaning revealed a problem and the system had nowhere to put it. That is Shine working as an inspection method and the standard failing to catch what it found — which is a real, specific, fixable finding, and it appears nowhere in the total.
Where to work next: Sort. Lowest score, earliest stage, and everything above it is resting on it. Removing the pallet jacks takes an afternoon, and every subsequent stage gets easier.
Dataset: ds-5s-goods-in-audit — the same data loads in the tool below, so you can reproduce every figure here yourself.
Your turn
The audit opens with the goods-in scores.
- Read the four findings. Two are about the dependency between stages — Sustain and Standardise being ahead of the basics — and those are the ones a paper form cannot make.
- Note what the tool shows large: the weakest stage. The total is deliberately small, because it is the number that tells you least.
- Set all five stages to 3. The uniform-score finding fires: five identical scores are almost always one overall impression written five times.
- Now score a healthy area — 4, 4, 3, 3, 2 — and watch every dependency finding clear. That shape is what a working 5S looks like, and it is upside down from the one people expect.
- Delete an observation and watch the score-without-evidence finding return. The sentence is the audit; the number is only a handle for it.
5S Audit
practiceWork on this next
Sort (seiri)
Total
11 / 20
Compare with this area last month — never with another area.
score without evidenceSet in order (seiton) has a score and no observation. A number on its own is an impression: it cannot be compared with next month's, and it tells the area nothing about what to change. Write the sentence — what you saw, and where.
score without evidenceSustain (shitsuke) has a score and no observation. A number on its own is an impression: it cannot be compared with next month's, and it tells the area nothing about what to change. Write the sentence — what you saw, and where.
sustain ahead of the restSustain is scored 4 while one of the first three stages is at 1. Sustain keeps the earlier stages in place; it cannot be in better shape than the things it holds. A high Sustain score over weak foundations usually describes a reliable audit routine rather than a workplace that stays that way.
standardise ahead of the basicsStandardise is scored 3 against a first-three low of 1. Standardise documents what Sort, Set and Shine established — a standard for a condition that does not exist is a document, and documenting it does not create it.
How this is calculated
Each stage is scored 0 to 4 — an even scale, so there is no comfortable middle. The total is shown small on purpose: a 5S score compares an area with itself over time, never with another area, because the scale is anchored by what a particular auditor has seen.
- Every stage is scored, and every score has an observation attached.
- Sustain is not more than one point ahead of the weakest of Sort, Set and Shine — it holds those in place and cannot be in better shape than the things it holds.
- Standardise is not far ahead of them either: it documents a condition, and documenting one does not create it.
- Five identical scores are flagged — that is one impression written five times.
Source: Hirano, H. (1995), 5 Pillars of the Visual Workplace; Osada, T. (1991), The 5S’s: Five Keys to a Total Quality Environment.
Check yourself
No hints. Wrong answers are explained, not softened.
An area scores Sort 2, Set 2, Shine 2, Standardise 4, Sustain 4. What does the shape tell you?
A site ranks its eight areas by 5S score in a monthly league table. What is the predictable consequence?
During Shine, an operator cleaning a machine finds a small oil leak. What does this tell you about the 5S programme?
Worth remembering
Why can Sustain not score higher than Sort, Set and Shine?
Sustain holds those stages in place; it cannot be in better shape than the things it holds. A high Sustain score over weak foundations describes a reliable audit routine, not a workplace.
What distinguishes Shine from tidying up?
Shine is cleaning AS inspection — hands on the equipment is how you find the leak, the wear, the missing fastener. If cleaning never reveals problems, it is housekeeping.
What may a 5S total be compared with?
The same area, last time. Never another area — the scale is anchored by what a particular auditor has seen, and a league table makes the score worth managing.
Can you do this now?
Rate yourself honestly. We compare your rating with how you actually answered — the gap is more useful than either number alone.
I can conduct a 5S audit of a workspace and score it against each of the five stages.
Your rating is recorded alongside your drill results. Neither alone marks the competency as met.