Building a shop-floor dashboard that answers one real question, not everything at once

This page explains how to build a shop-floor dashboard around one real question a factory actually needs answered, instead of a screen full of every figure a system can produce

This page is for a merchandiser or production manager at a garment factory who already has shop-floor data, from a paper log, a machine feed, or a spreadsheet, but has never turned that data into a dashboard someone actually uses to decide something. At work, a person handed a screen of ten charts and no single figure that drives a decision looks at it once, then goes back to walking the floor and asking the line supervisor directly. This lesson mainly teaches the information flow: it shows how to pick the one real question a dashboard must answer, before choosing what to put on the screen.

Northline Apparel's buying office asks Delta Garments what its own shop-floor dashboard is actually for on Order A, a Make to Order export order of 15,000 T-shirts with a fixed port cut-off on day 13 of the 16-week timeline. Delta Garments' production team already tracks output, downtime, and defects on the line producing Order A. Without deciding which figure actually answers Northline's real concern, whether the order ships on time, the dashboard is just data on a screen, not a tool anyone acts on.

You need: no previous lesson from this track. A working knowledge of how a shop-floor line records its own daily output is assumed but not required to follow this lesson.

By the end of this lesson you can:

• name the one real question a shop-floor dashboard must answer before choosing what to display;
• tell a vanity metric apart from a figure that actually drives a decision;
• set a dashboard's refresh cadence and threshold to match the decision it supports;
• assign an owner responsible for acting on the dashboard's own figure;
• recognize when a factory has built a dashboard around convenience instead of around a real question.

Not covered here: calculating a line's own efficiency figure, covered in lesson T26-C03-L02, and choosing which shop-floor metric predicts a late shipment, covered in lesson T26-C03-L03.

Words you will meet in this lesson

A shop-floor dashboard is only useful once it is built to answer one specific, real decision, not to display every figure a system can produce

A dashboard's own value comes from the decision it supports, not from the number of charts on its screen. The general rule is to name the one real question first, such as whether an order will ship on time, and then choose only the figures, refresh cadence, and threshold that answer that one question, before anything else is added to the screen.

The one real question a dashboard answers changes with the order, the model, and who is asking

Northline Apparel's real question for Order A is whether sewing finishes before the port cut-off

Delta Garments already records output, downtime, and defects for the line producing Order A. Mona, the merchandiser, sits with Karim, the production manager, to decide what the line's own dashboard should actually answer for Northline Apparel. Order A's 15,000 pieces must clear sewing within its own 2-week window inside the 16-week timeline, so that shipping documents and the port cut-off on day 13 are still reachable.

Turning Order A's own timeline into the one daily figure the dashboard actually needs

Once the sewing window is turned into a single required daily output, the dashboard's one real question becomes simple to state: is today's actual output at or above 1,500 pieces, and by how much does it fall short if it is not. Every other figure the line already records, downtime causes, defect types, individual operator counts, stays available for a drill-down, but none of them replaces this one number on the main screen.

Five steps a factory follows to build a dashboard around one real question

1. Name the one decision the dashboard must support, stated as a plain question before any figure is chosen.
2. Calculate the single figure, or comparison, that answers that question directly, as shown above for Order A's own required daily output.
3. Set the refresh cadence to match how fast the decision actually needs to be made, hourly for a tight sewing window, daily for a longer one.
4. Choose the alert threshold that tells the dashboard owner when the figure has crossed from acceptable to a problem worth acting on.
5. Assign a dashboard owner by name, responsible for checking the figure against its threshold and acting on it.

A common mistake is filling a dashboard's main screen with every figure a system can produce, so the one figure that answers the real question gets lost

A common mistake is a factory building a dashboard by asking what a system can display rather than what decision it needs to support, so the main screen ends up with ten charts, output, downtime, defects, attendance, and machine speed, all shown with equal weight and none of them named as the figure that actually matters most today. What this costs is real: Mona and Karim can spend several minutes each morning scanning a crowded screen before finding whether Order A is actually on pace, time a single, correctly chosen figure would have saved, and a delay that, on a 2-week sewing window against a fixed port cut-off, a factory cannot easily recover once lost.

• A dashboard's one real question should be named before any figure is chosen for its screen.
• A vanity metric looks informative but does not change what anyone decides or does next.
• Refresh cadence should match how fast the underlying decision actually needs to be made.
• An alert threshold, not a raw figure alone, is what tells a dashboard owner when to act.
• A hybrid line running two styles needs its dashboard to separate their own questions, not blend them into one figure.
• Every dashboard needs a named owner responsible for acting on its own figure.

Karim wants to add six more charts to Order A's own dashboard because Delta Garments' system can already produce them

Karim, the production manager, has just learned that the factory's shop-floor system can also chart individual operator speed, thread consumption, and machine idle time for the line running Order A. He suggests adding all three, alongside the three charts already on the dashboard, so the screen shows "everything available" before Northline Apparel's next check-in call.

Should Karim add all three new charts to Order A's own dashboard before the buyer call? Decide before opening the answer.

Dashboards and shop-floor analytics worksheet: state the one real question and its own figure

Open the course workbook's "One real question" sheet. Enter the decision the dashboard must support, the single figure that answers it, the refresh cadence, the alert threshold, and the dashboard owner's name; the sheet checks that every field needed to build a working dashboard is actually filled in.

[Download the Dashboards and shop-floor analytics worksheet](../../templates/T26-C03-dashboards-and-shop-floor-analytics-worksheet.xlsx)

Download the Dashboards and shop-floor analytics worksheet (Excel)

Lesson 3.2: Calculating a line's own efficiency from real, not assumed, output data

You now know how to build a dashboard around one real question. The next lesson shows how to calculate the actual figure, a line's own efficiency, that a good dashboard puts in front of that question. Back to the course map