What to Track on a Shop Floor Dashboard (And What to Leave Off)

Most shop floor dashboards fail. Not because they're missing data. Because they show too much of the wrong data.

Walk into any manufacturing facility and you'll see it: a TV mounted on the wall, displaying a dashboard nobody looks at. Rows of numbers. Charts from last week. Metrics that seemed important when someone built the thing six months ago.

That's not a dashboard. That's expensive wallpaper.

The difference between a dashboard that drives action and one that fades into the background comes down to one question: does it help the person looking at it make a better decision right now?

Here's how to build a shop floor dashboard that actually gets used.

Why Most Shop Floor Dashboards Fail

The first mistake is trying to show everything. If you have the data, the thinking goes, you should display the data. Wrong.

More metrics means more noise. When everything is highlighted, nothing is. Operators scanning a screen with 20 different numbers will default to ignoring all of them.

The second mistake is building dashboards by committee. IT pulls what's easy to extract. Finance wants their numbers. Quality wants theirs. The result is a Frankenstein that serves no one well.

The third mistake is update speed. A dashboard showing 15-minute-old data isn't showing reality. It's showing history. By the time you react, the situation has already changed.

Dashboards fail when they're designed to display data instead of enable decisions.

The Three Questions Every Metric Must Answer

Before you add any metric to your shop floor dashboard, it must answer at least one of these questions:

Is something wrong right now?

This is the most critical function. If a machine stops, if quality falls off a cliff, if a safety issue arises, the dashboard should make it obvious immediately.

Are we on track for today?

Operators and supervisors need to know whether they're hitting targets or falling behind. Not last week's targets. Today's. Real-time progress toward the shift goal.

Is there a trend I need to address?

Some problems don't show up as sudden spikes. They creep in gradually. A dashboard should surface trends before they become crises.

If a metric doesn't answer one of these questions for the person looking at it, it doesn't belong on the shop floor. Move it to a report. Put it in a monthly review. Just don't clutter the real-time view.

Essential Metrics for Operators

Operators don't need comprehensive analytics. They need to know what's happening right now and what they should do about it.

Tier 1: Right Now

Machine status. Is it running, stopped, or idle? This sounds basic, but it's the foundation. Green means running. Red means stopped. Yellow means attention needed. No interpretation required.

Current job progress. How many pieces are complete versus the target? If the order calls for 500 parts and you're at 340, you know exactly where you stand. Show it as a simple progress bar or count.

Active alerts and alarms. Anything that requires immediate attention gets priority placement. Don't bury alarms in a list. Make them impossible to miss.

Tier 2: Today's Picture

Daily production target vs. actual. The shift target is 1,000 parts. You're at 750 with two hours left. Are you on pace? Behind? Simple math, but operators shouldn't have to calculate it themselves.

Quality count. Scrap and rework for the current shift. If defects are piling up, operators need to see it in real time, not discover it at end-of-day review.

Next job in queue. What's coming after this order finishes? Knowing the next job helps operators prepare, identify potential changeover issues, and avoid idle time.

Tier 3: Quick Reference

Current work order details. Part number, customer name, special instructions. Information operators need to reference without leaving the station.

Material availability. Are the materials for the next job staged and ready? Flagging shortages before they cause downtime beats discovering them mid-changeover.

Essential Metrics for Managers

Managers aren't operating machines. They're making decisions about resources, priorities, and problems. Their dashboard needs different information.

Tier 1: Floor Status at a Glance

Overall Equipment Effectiveness (OEE). OEE combines availability, performance, and quality into a single percentage. An OEE of 85% tells you that 85% of your potential productive time is actually producing good parts. It's a useful snapshot, but it's a lagging indicator. For a deeper dive, see our guide to OEE.

Schedule adherence. Are you on time, ahead, or behind on today's orders? A color-coded status for each major job or work center gives managers a quick read without drilling into details.

Bottleneck visibility. Where is work piling up? If the CNC cell is drowning while assembly sits idle, that imbalance needs to surface immediately.

Tier 2: Performance Trends

Daily and weekly output trends. Is throughput steady, improving, or declining? A simple line chart over the past 5-10 days reveals patterns that single-day numbers hide.

Quality trends. Is your defect rate stable or creeping up? Spotting a gradual increase early prevents bigger problems downstream.

Lead time tracking. How long are jobs actually taking versus estimated? Consistent overruns signal estimation problems or hidden inefficiencies.

Tier 3: Resource View

Labor utilization by area. Which cells are fully staffed and which are running short? Resource allocation decisions need current headcount visibility.

Machine utilization by cell. Which equipment is running hot and which has capacity? This drives decisions about overtime, outsourcing, and scheduling.

Inventory levels for critical items. Stock levels for materials that frequently cause delays. Not your entire warehouse, just the items that tend to stop production when they run out.

Metrics That Waste Dashboard Space

Knowing what to include is half the battle. Knowing what to leave off is the other half.

Anything updated less than hourly. If a metric only refreshes once a day, it belongs in a report, not a real-time dashboard. Stale data trains people to ignore the screen.

Metrics nobody acts on. If showing a number never prompts anyone to do anything differently, why is it there? Every metric should connect to a possible action.

Vanity metrics. Cumulative totals without context tell you nothing useful. "We've made 50,000 parts this year" sounds impressive but doesn't help anyone make a decision today.

Financial data. Revenue, cost per unit, margin percentages. These matter for executives, not shop floor operators. Wrong audience, wrong dashboard.

Excessive precision. Showing OEE as 83.7429% implies false accuracy and makes the number harder to read. Round to 84% and move on.

The goal is a dashboard you can understand in five seconds, not a data dump that requires analysis.

Real-Time vs. Batch: Why Update Speed Matters

There's a meaningful difference between real-time data and near-real-time data. It's not just semantics.

A dashboard that updates every 15 minutes shows you where you were 15 minutes ago. In that time, a machine could have stopped, a quality issue could have emerged, or an operator could have already fixed the problem you're just now seeing.

Real-time means updates happen as events occur. When a machine status changes, the dashboard reflects it within seconds, not minutes.

The technology behind this matters. Traditional systems use polling: they ask the database "what changed?" on a schedule. Modern systems use WebSocket connections that push updates the instant something happens. The difference in responsiveness is significant.

For operators, real-time visibility means they can react before small problems become big ones. For managers, it means the dashboard actually reflects current reality when they walk the floor. For everyone, it means the dashboard stays relevant instead of becoming background noise.

If you're evaluating shop floor visibility solutions, update architecture should be a key criterion. Dashboards that update in seconds serve a different purpose than dashboards that update in minutes.

Building a Dashboard That Gets Used

The best dashboard is the one that prompts action. If operators check it habitually, if supervisors reference it in shift handoffs, if managers use it to ask better questions, it's working.

Start simple. Begin with five or six metrics that directly answer the three questions: what's wrong, are we on track, what's trending. Add more only when someone has a specific use case for the new information. Resist the urge to add "nice to have" metrics. Every addition dilutes the signal.

Design for the audience. What operators need differs from what plant managers need. One screen won't serve both roles well. Consider role-based views that show relevant information to each user. An operator terminal at the work center should show different information than the summary view in the plant manager's office.

Prioritize visibility over density. A dashboard with ten clearly displayed metrics beats one with fifty crammed into every corner. White space isn't waste. It's clarity. If someone has to squint or search, the dashboard fails its purpose.

Test with real users. Watch operators interact with the dashboard. Ask what's confusing. Ask what's missing. The people using it daily will tell you what the design missed. The best feedback comes from the floor, not from conference room reviews.

Place dashboards where decisions happen. A great dashboard in a break room helps no one. Put screens where operators can see them from their stations. Put manager views in offices or on tablets they carry during floor walks.

Finally, measure whether it's working. Are operators checking it? Do supervisors reference it in shift handoffs? If people still rely on walking the floor or calling around for status updates, the dashboard isn't doing its job. The goal is to replace tribal knowledge with visible information.

Ready to see your shop floor in real time? Book a demo and we'll show you what Workcell looks like with your actual production data.