Concept

What to Put on a Wallboard Operators Actually Look At

Most shop-floor TVs get ignored within a week. What earns a second glance, what's decoration, and how to avoid a wall of red nobody trusts.

9 min read · Last reviewed August 19, 2026

Walk any shop floor with a TV mounted on the wall and there’s a decent chance nobody’s actually looking at it, not because the operators don’t care about the numbers, because the board stopped saying anything useful somewhere around the second week. A wallboard that gets ignored isn’t a display problem, it’s usually a content problem, and it’s worth being specific about what earns a glance and what’s just there because someone thought a TV needed filling.

The test that actually matters: would someone change what they’re doing because of this

Every tile on a wallboard should pass one test. If this number moved, would anyone on the floor do something differently in the next ten minutes. A live OEE percentage passes, a bad trend is a reason to walk over and ask what’s wrong. A part count against target passes, it tells an operator whether they’re behind before the end of the shift catches them by surprise. A decorative gauge that always sits comfortably in the green passes nothing, it’s wallpaper, and wallpaper is exactly what people stop seeing.

This is also why a board built once and never revisited tends to go stale. The right tiles for a line that’s fighting a changeover problem this month aren’t the same tiles that mattered last quarter when the issue was scrap on one specific part. A wallboard that gets checked every few months against what’s actually being managed right now stays useful. One that gets set up at go-live and never touched again slowly turns into background noise.

Would someone act on this tile OEE, trending down Someone walks over. Asks what changed. Passes the test. Gauge, always green Nobody's ever acted on it. Never will. Wallpaper.
The test isn't "is it accurate," most decorative tiles are. The test is whether it ever changes what someone does next.

No color that doesn’t mean something

The fastest way to train people to stop trusting a wallboard is decorative color, a tile with a cheerful green accent that isn’t reporting anything, just sitting there looking healthy. Once someone notices that green doesn’t actually mean “good” on that particular tile, they stop trusting green anywhere on the board, and the whole signal value of color collapses. A tile’s accent should be quiet by default and only turn green, amber, or red when the metric it’s showing actually enters that state. No decorative color, no exceptions, that discipline is what keeps color meaning something the one time it actually needs to.

Read it from across the room, not from your desk

A wallboard is read from fifteen feet away by someone walking past, not studied up close on a monitor. That changes what belongs on it more than most people expect going in. Dense tables with six columns of small text are exactly the kind of thing that looks informative on a laptop and reads as noise on a TV, nobody’s going to squint at row four from across an aisle. Fewer numbers, bigger, wins every time over more numbers, smaller. If a tile needs a legend to explain what it’s showing, it’s already failed the from-across-the-room test.

Language and access are part of “actually looked at” too

A board nobody can read isn’t ignored, it’s useless from the start, and that’s a more literal problem than it sounds on a floor with a mixed workforce. If a meaningful share of the people walking past speak a different first language than whoever built the board, a language toggle on the board itself is worth having rather than assuming everyone reads it the same way the office does. It’s a small thing until it isn’t, a board that only speaks one language to a bilingual floor excludes exactly the people it’s meant to inform. The same logic applies to physical placement, a board mounted where it’s actually in someone’s daily path gets glanced at. One tucked in a break room corner mostly gets read during breaks, which is a different, narrower audience than the one a shop-floor board is usually built for.

Live data or a visible gap, not a stale number pretending to be live

The one thing worse than a board with too little on it is a board that’s wrong without anyone knowing. A tile showing yesterday’s number as if it were right now is worse than no tile at all, because it actively misleads someone who glances up mid-shift expecting a current answer. If a machine’s telemetry stops flowing, the board should say so plainly, a clear no-data state, dashed values, whatever makes the gap obvious, rather than freezing on the last real number and letting it go stale in place. A visible gap earns trust back the next time the board recovers. A stale number that looked live burns trust the moment someone notices, and once burned, that’s the last time they check the board first.

Build it for the line it’s actually on, not a generic template

The right mix of tiles depends on what’s actually being managed on that line, not a one-size template copied across every board in the building. A cell fighting changeover time wants a changeover or downtime-reason tile front and center, that’s the number someone should be watching to know whether this week’s push is working. A line where the real risk is a quality escape wants scrap rate and first-pass yield in easy view instead. A flow-tracked assembly cell benefits from live per-edge rate on a facility map style tile, showing where in the line things are actually backing up right now, instead of a single aggregate number for the whole cell. Build the board around the question that line actually needs answered this month, and revisit it when that question changes.

A board with nothing to say should say nothing loudly

Not every line needs six tiles. A quiet, well-running cell with one clean OEE number and a downtime Pareto underneath it is a better board than a busy one crammed with tiles nobody asked for just because there was room. If a tile’s real answer is “nothing’s wrong here,” let it say that plainly and take up less space, rather than padding the board to look thorough. A wallboard’s job is to be worth a glance, not to look impressively full.

Quick recap

  • Every tile should pass one test: if this number moved, would someone do something differently
  • No decorative color, an accent that isn’t reporting anything trains people to stop trusting color anywhere on the board
  • Design for fifteen feet away, fewer bigger numbers beat dense tables every time
  • A gap in the data should look like a gap, not a stale number frozen in place and passed off as current
  • Build the tile mix around what that specific line is actually fighting this month, not a copy-pasted template
  • A quiet board with one real number beats a busy board full of tiles nobody asked for

Related guides

From the Help Center

$200/machine/mo · pilots from $4,500 · hardware included

See full pricing →

Bring your machine list. We'll tell you exactly what plugs in.

30 days, hardware included, line pilot $4,500 or plant pilot $8,500, fully credited when you expand.