Visual Management: Tools and Examples for the Workplace
Turn this article into takeaways for your work.
Each assistant summarizes the article only for you and suggests best practices for your work.
Ask five people how a piece of work is going and you'll get five answers, all sincere and at least two of them wrong. The status is real. It just lives somewhere nobody can see it: in one person's head, in a spreadsheet opened on Thursdays, in a thread three apps away, or in a meeting that hasn't happened yet.
Visual management pulls that status into the open, into the space where the work happens, so anyone walking past can tell in seconds whether things are fine or not fine. It isn't decoration and it isn't reporting. Reporting tells you what happened last month. Visual management tells you what's happening now, without anyone having to ask.
Key Facts: Visual Management
- The Lean Enterprise Institute defines visual management as "the placement in plain view of all tools, parts, production activities, and indicators of production system performance, so the status of the system can be understood at a glance by everyone involved" (LEI Lexicon).
- A 2025 peer-reviewed taxonomy calls visual management "a sensory system designed to provide information that guides people's actions", and credits Gwendolyn Galsworth (1997) with the four-device split into indicator, signal, control and guarantee (Construction Management and Economics, 2025).
- People pick a named target picture out of a rapid image stream at above-chance rates even when each image shows for only 13 milliseconds, one reason a well-placed visual signals state faster than a paragraph (Potter et al., Attention, Perception, & Psychophysics, 2014).
- A 2024 systematic review of healthcare performance visualisation found 10 of its 12 studies reported a statistically significant change in at least one outcome, while concluding only that such tools "have the potential to improve clinical performance indicators" (BMJ Open, 2024).
- On a Toyota line, "when equipment stops, the andon (problem display board) lights up to notify workers", and workers can pull a stop cord to call the person in charge (Toyota).
What Visual Management Actually Is
The Lean Enterprise Institute's definition is worth reading twice, because the last five words carry the weight: the status of the system can be understood at a glance by everyone involved (LEI). Not by the supervisor, not by whoever built the spreadsheet. By everyone, including whoever joined last week.
A 2025 paper frames it more broadly, as "a sensory system designed to provide information that guides people's actions", built from devices that "work together through shared conventions tailored to the context" (Construction Management and Economics, 2025). Shared conventions is the part teams skip. A red magnet means nothing until everyone agrees what red means and what happens next.
A working definition: visual management makes the standard, the current state, and the gap between them visible where work happens, so deviations announce themselves.
Where this page draws the line against 5S, gemba walks and andon
Four ideas in this collection sit close together, and treating them as synonyms is the fastest way to build something that looks lean and does nothing.
5S is workplace organization: sort, set in order, shine, standardize, sustain. It decides what stays, where it lives and how clean it's kept, using the red-tag process to clear out what doesn't belong. 5S produces a workplace that can be read visually. It's the groundwork, not the reading.
A gemba walk is the act of going to see. A leader leaves the desk, stands where the work happens, and observes. It's a behaviour on a cadence, and it depends on someone showing up.
Andon is one instrument: the signal-and-escalation system. Something goes wrong, a light or board announces it, and a named person responds inside a defined window. Andon is visual management with a response contract attached.
Visual management is the umbrella over all three, the discipline of making the state of work legible at a glance so a deviation is obvious without anyone asking. 5S makes the space readable. The gemba walk is a person reading it. Andon is the loudest instrument in the kit. This page covers the kit, the principle behind it, and how to build one that survives past month three.
The Problem It Solves
Every operation has a status somewhere. The only question is how expensive it is to retrieve.
In an unmanaged team, retrieving status costs a message, a wait, a context switch for whoever answers, and often a meeting whose whole purpose is moving information from private storage into shared storage. That meeting creates nothing a customer would pay for. It's muda: waiting and over-processing.
Three patterns show up over and over:
Status lives in someone's head. Only Priya knows which of the twelve open orders are at risk. When Priya is on leave, the team is blind. That's a single point of failure disguised as expertise.
Status lives in a system nobody opens. The data is correct, current and functionally invisible, because reading it requires deciding to go look. Information you have to seek gets consulted after a problem, not before one.
Status only exists at meeting time. The team finds out on Tuesday that Thursday's commitment is already impossible. It was knowable on Friday. That delay between a problem becoming true and becoming visible is the real cost, and it compounds.
Visual management attacks the retrieval cost directly. Put the board where people already walk or look, and status stops being something you have to request.
The Principle: Standard, Then Deviation, Then Response
Underneath every good visual control is the same three-part chain, and it only works in order.
Make the standard visible. A number alone is inert. "47 tickets open" tells you nothing until you know whether 47 is normal. Post the target beside the actual: the takt rate, the SLA commitment, the WIP limit, the shelf position an item belongs in. Standard work is the reference the visual is measured against, which is why teams without documented standards watch their boards become decoration.
Make the deviation obvious against it. Once a standard exists, the visual's job is to make the gap jump out. A shadow board does this beautifully: an empty outline is a missing tool, no counting required. A pace board plots planned against actual cumulatively, so falling behind shows as a widening gap rather than a number to interpret.
Make the response obvious. This separates working visual management from wall art. When the visual goes red, who does what, and by when? On a Toyota line the answer is built in: the cord gets pulled, the andon lights, the team leader comes. As with jidoka, the response is defined in advance so nobody invents one under pressure. A signal with no defined response teaches people that red is normal, and a red that means nothing is worse than no colour at all.
The Four Levels of Visual Control
This is the spine of the subject, and it explains most failed implementations in one move. Gwendolyn Galsworth, in Visual Workplace, Visual Thinking, sorts visual devices into four ascending power levels: indicator, signal, control, and guarantee. The 2025 taxonomy paper credits the split to her 1997 work (Construction Management and Economics). The levels differ by how much they change behaviour, not how much information they carry.
| Level | What it does | Compliance | Factory example | Office example |
|---|---|---|---|---|
| Visual indicator | Shows information. Acting on it is voluntary. | Optional | A bin label, a posted shift target, yesterday's output on a whiteboard | A dashboard tile, a posted OKR, a shared status doc |
| Visual signal | Grabs attention when something changes. Still voluntary, but hard to ignore. | Strongly prompted | An andon light, a flashing beacon, a red card on a rack | A red column on a task board, an alert when a queue breaches its SLA |
| Visual control | Constrains behaviour through physical or structural limits. | Enforced by structure | A floor square holding exactly six pallets, a kanban card count | A WIP limit that blocks a card entering a column, a form that won't submit incomplete |
| Visual guarantee | Makes the wrong action impossible. | Absolute | A fixture accepting the part in one orientation only, a connector that won't mate backwards | A workflow that can't advance without approval, a field rejecting invalid input |
Most teams build indicators, then wonder why nothing changes. An indicator informs someone who was already going to do the right thing. It does nothing about the person who is busy, distracted, or working around the standard on purpose. To change behaviour, climb the ladder.
The top rung has a name in lean vocabulary: a visual guarantee is poka-yoke, error proofing, where the design of the work makes a mistake impossible rather than merely discouraged. It's the most reliable and the most expensive to build. So the honest pattern is indicators for context, signals for exceptions, controls for rules you care about, guarantees for failures you cannot afford.
A quick audit: count your devices by level. If more than three-quarters are indicators, you have a display, not a management system.
A Working Catalogue of Visual Management Tools
Each of these does one job. Choosing the wrong one is the most common design error after stopping at indicators.
| Tool | What it's genuinely for | Level | Fails when |
|---|---|---|---|
| Kanban / task board | Showing where each item sits in a flow, and capping work in progress | Control with WIP limits, indicator without | Columns don't match how work moves, or no limits exist |
| Shadow board | Making a missing tool visible via its outlined home position | Control | Outlines exist for tools nobody uses, so gaps stop meaning anything |
| Floor and lane marking | Defining where material, traffic and people belong | Control | Painted for tidiness, not a rule, so crossings go unnoticed |
| Andon light or board | Calling a named responder to an abnormality inside a set window | Signal | Nobody responds, or it informs rather than escalates |
| Production or pace board | Comparing planned against actual cumulatively through the shift | Indicator or signal | Updated at shift end, turning a signal into a post-mortem |
| Standard work at the point of use | Putting the best-known method where the work is done | Indicator | It's out of date, so people learn to ignore it |
| Heijunka box | Levelling mix and volume by slotting cards into time intervals | Control | Used as a schedule display, not the thing releasing work |
| Colour coding | Encoding a category or state so it's readable without reading | Indicator or signal | Too many colours, or colours meaning different things by area |
| Digital dashboard or wallboard | Surfacing automated data a human couldn't update by hand | Indicator | It shows what's easy to measure, not what drives a decision |
| Daily management board | Anchoring a short standing huddle on status, blockers and owners | Signal, with a response rule | It becomes a status recital with no decisions attached |
Four of these deserve a note.
Kanban is a signal before it's a board. The Lean Enterprise Institute defines a kanban as "a signaling device that gives authorization and instructions for the production or withdrawal (conveyance) of items in a pull system" (LEI). The card count is the control. The board is how you see it. Copy the board without the limit and you get a prettier to-do list.
Andon is defined by its response, not its bulb. LEI calls it "a visual management tool that highlights the status of operations in an area at a single glance and that signals whenever an abnormality occurs" (LEI). Toyota makes the loop explicit: the board lights up to notify workers, and workers pull the stop cord to call the person in charge (Toyota).
Heijunka is a control, not a chart. Heijunka is "leveling the type and quantity of production over a fixed period of time" (LEI), and the box does that by holding cards in physical time slots. If work is released from somewhere else, the box is decoration.
Floor marking is often a legal requirement. OSHA requires that "permanent aisles and passageways shall be appropriately marked" (29 CFR 1910.176(a)), and the safety colour code reserves red for danger, fire protection equipment and emergency stops, yellow for caution and physical hazards (29 CFR 1910.144). Choose board colours knowing red and yellow already carry meanings.
Visual Management in Office and Knowledge Work
Knowledge work needs this more than a factory does, because its inventory is invisible. A pile of unfinished parts is obvious from across the room. Forty half-finished tickets look exactly like none. The translation is direct once you stop looking for machines.
| Factory device | Office or knowledge-work equivalent | What it makes visible |
|---|---|---|
| Production pace board | A committed-against-delivered chart for the week | Drift behind commitment while there's time to act |
| Andon light | An alert routed to a named owner when a queue breaches its target | An abnormality needing a human now, not at stand-up |
| Kanban card count | A WIP limit enforced by the board, so a full column refuses cards | Overload, before it becomes multitasking and rework |
| Shadow board | A checklist inside the work item, where an empty box is a gap | A missing step in a handover |
| Floor marking | Ownership lanes in a swimlane diagram, or a column per team | Where work sits and who owns it now |
| Standard work at the station | The definition of done or runbook, linked from the ticket | The agreed method, at the moment of use |
| Heijunka box | A capped weekly intake splitting planned from interrupt work | Whether unplanned demand is eating the plan |
Two office-specific patterns are worth adopting deliberately.
Make the commitment visible next to the actual. Support teams already have a standard sitting in a contract. Putting the SLA target on the same board as current queue age turns an abstract number into a live signal, and gives red a meaning everyone accepts.
Make ageing visible, not just position. A card that has sat in review for nine days looks identical to one that arrived this morning unless you show age. Ageing is the office equivalent of inventory piling up, and it's what most task boards are missing.
Digital tools are fine here. The mistake is assuming the tool is the practice.
How to Design a Board People Actually Look At
Most boards die the same way: they start rich, get stale, and become wallpaper. These decisions separate the ones that survive.
| Design check | What good looks like | Failure signal |
|---|---|---|
| One owner | A named person is accountable, and the name is on the board | "We all keep it updated", meaning nobody does |
| One glance | An outsider tells fine from not-fine in five seconds | People must read numbers and do arithmetic first |
| One decision | Every element changes something someone does today | Metrics nobody has acted on stay posted |
| Updated by the doers | The team writes on it during the work, in their own tool | An analyst updates it the night before the review |
| Where the work happens | In the team's space, or on a screen they keep open | A board in a room the team enters weekly |
| Standard beside actual | Target, limit or commitment next to the current number | A raw number with no reference point |
| A defined response | Red carries a written rule: who is called, inside what window | Red sits there for weeks and nothing happens |
| Small enough to maintain | One surface, under five minutes a day | Half an hour, so it gets skipped, then abandoned |
The "one decision" test is the harsh one. Point at each element on an existing board and ask what someone does differently because of it. Whatever survives stays. Everything else comes off, and the board improves.
A caveat from the evidence belongs here. The 2024 review noted that none of its studies examined how clinicians actually used the displayed information, leaving "few insights into why some clinicians opted not to engage with the display information" (BMJ Open, 2024). Research measures the artefact. The checks above are about the practice around it, the harder and less studied half.
So pair the board with a short standing huddle in front of it. That huddle converts a display into continuous improvement: the board raises the deviation, the huddle assigns the response, tomorrow's board shows whether it worked.
Physical Boards vs Digital Dashboards
Digital dashboards fail more often than physical boards, and the reason isn't technical.
A physical board has friction, and friction turns out to be a feature. Someone has to stand up, walk over, and write. That's a daily commitment made in front of colleagues, and it makes neglect visible: a board untouched for four days looks abandoned. A dashboard nobody has opened in four days looks exactly like one watched closely. Automatic data removes the ritual, and with it the shared attention that made the board a practice rather than a feed.
Dashboards also drift toward what the system can measure rather than what the team must decide, because adding a chart costs nothing. That's how a screen ends up with fourteen tiles nobody acts on.
Digital genuinely wins in several cases.
| Situation | Better choice | Why |
|---|---|---|
| Team spread across sites or time zones | Digital | A physical board only serves people who walk past |
| Data changes faster than a human can write it | Digital | Manual updates can't keep pace, and stale numbers mislead |
| You need history, audit trail or trend analysis | Digital | Whiteboards don't remember, and photos aren't a data set |
| Numbers must reconcile with a system of record | Digital | Hand-copied figures drift, then people argue about figures |
| Co-located team doing daily execution work | Physical | Cheap, instantly editable, a place to stand together |
| The metric set is still being figured out | Physical | Changing a whiteboard takes a marker, a dashboard takes a ticket |
The practical rule: start physical if the team is in one place, and digitize a visual only once it has earned its place by driving a decision for a month. That order also protects you from dashboard sprawl, because everything on the screen survived a manual phase first.
Measuring Whether It's Working
Visual management is easy to declare successful and hard to prove, so measure the behaviour it should change rather than the artefact.
Time to detect. How long between a problem becoming true and someone knowing? Sample recent issues and ask when each first became knowable. This is the number visual management attacks most directly.
Response rate against signals. How often did a red state get its defined response inside the window? Many signals with few responses means an alarm nobody answers.
Meeting time spent on status. A working board shortens the standing meeting, because status has moved to the wall. A huddle still spending twenty minutes on "where are we" means the board isn't carrying its load.
Whether a newcomer can read the area. Give an outsider thirty seconds and ask if the area is ahead or behind. That's the closest thing to a direct test of the LEI definition.
Freshness. Track the age of the last update. Staleness is the leading indicator of a board about to die.
Pair these with your process KPIs and the routines you use to monitor the process. Visual management isn't a separate measurement system. It's a delivery mechanism for measures you already track.
Where Visual Management Goes Wrong
The failure modes are worth naming, because most of them look like success for the first few weeks.
| Failure mode | What it looks like | The fix |
|---|---|---|
| Board theatre | Immaculate boards built for a visitor tour, untouched by the team | Let the team design it, in their handwriting, for their use |
| All indicators, no controls | Plenty of posted information, no change in behaviour | Climb the ladder: add limits, owned signals, error proofing |
| Dashboard sprawl | Fourteen tiles, no decisions, nobody can say which matters | Delete anything that hasn't driven an action in a month |
| Vanity metrics | Numbers that only go up and always look good | Show a target and a gap, and keep one measure that can look bad |
| Red with no response | The signal fires regularly and nothing happens | Write the response rule first, or skip the signal |
| Stale by design | Updating depends on someone with other priorities | Shrink the board until updating takes five minutes |
| Copying the artefact | A heijunka box that doesn't release work, a board with no WIP limit | Ask what rule it enforces. No rule means you have a picture |
One limit is worth stating plainly: visual management makes problems visible, and that's all it does. Make deviations obvious in an organization with no capacity to respond and you've built a machine for generating visible, unresolved problems, which erodes trust faster than the opacity did. Before adding signals, check someone can act on them.
Visibility isn't diagnosis either. A board tells you the queue is backing up. It won't tell you why, which is what a value stream map or a walk to the gemba is for. Visual management sits inside lean methodology as the sensing layer, and earns its keep when the improvement routines above it already run.
Frequently Asked Questions about Visual Management
What is visual management in simple terms?
It's making the state of work visible where the work happens, so anyone can tell at a glance whether things are on track. The Lean Enterprise Institute defines it as placing tools, activities and performance indicators in plain view "so the status of the system can be understood at a glance by everyone involved". A deviation announces itself instead of waiting to be found in a meeting.
What's the difference between visual management and 5S?
5S is workplace organization: sorting, setting in order, cleaning, standardizing and sustaining. Visual management is the broader discipline of making status and standards legible at a glance. 5S creates a workplace that can be read visually, and visual management is what you build on top of it.
Is an andon the same thing as visual management?
No. Andon is one instrument within visual management: a signal-and-escalation system that announces an abnormality and calls a named responder inside a defined window. Visual management is the umbrella that also covers boards, floor marking, shadow boards, standard work at the point of use, and error proofing.
What are the four levels of visual control?
Indicator, signal, control and guarantee, a ladder credited to Gwendolyn Galsworth. An indicator only informs, a signal grabs attention when something changes, a control constrains behaviour such as a work-in-progress limit, and a guarantee makes the wrong action impossible. Most teams build only indicators, which is why their visuals rarely change what people do.
Does visual management work in an office or only in manufacturing?
It works in offices, agencies and software teams, and arguably matters more there, because knowledge work has invisible inventory. Task boards with work-in-progress limits, ageing indicators, SLA targets beside live queue age, and a definition of done on each item are direct office equivalents of shop-floor devices.
Are digital dashboards better than physical boards?
Not by default. Physical boards create a daily ritual and make neglect obvious, which is often what makes the practice stick. Digital wins when the team is distributed, when data changes faster than a human can write it, when you need history and an audit trail, or when figures must reconcile with a record system.
How do I know if our visual management is actually working?
Measure behaviour, not artefacts. Track how long a problem takes to become visible, how often a red state got its defined response inside the window, and whether a newcomer can tell in thirty seconds if the area is ahead or behind. Watch the age of the last update: staleness predicts a board's death long before anyone admits it.
Visual management isn't a board or a dashboard. It's a commitment to make the standard visible, the gap against it obvious, and the response defined in advance. Get those three in order and almost any artefact works. Skip the third and you've built a very tidy way of watching problems go unfixed.
Senior Operations & Growth Strategist
On this page
- What Visual Management Actually Is
- Where this page draws the line against 5S, gemba walks and andon
- The Problem It Solves
- The Principle: Standard, Then Deviation, Then Response
- The Four Levels of Visual Control
- A Working Catalogue of Visual Management Tools
- Visual Management in Office and Knowledge Work
- How to Design a Board People Actually Look At
- Physical Boards vs Digital Dashboards
- Measuring Whether It's Working
- Where Visual Management Goes Wrong