How to Measure Response Time on the Production Floor (Without New Hardware)

Ask a supervisor how long it takes maintenance to respond to a stoppage and you'll get an answer. Ask how they know, and the honest answer is usually "it feels like." That's not a criticism. Most plants simply have no record of it.

Response time is one of the most useful numbers on a production floor, and one of the least measured. Here's a practical way to start.

Define what you're measuring

Before anyone writes anything down, agree on three moments.

Requested: when the operator asked for help.

Claimed: when a specific person said "I've got it."

Completed: when the problem was resolved and the operator could continue.

That gives you two numbers that tell different stories. Requested to claimed is response time: how long a request sits before someone owns it. Claimed to completed is fix time: how long the work takes once someone is on it. Plants are often surprised to find the first one is the bigger problem.

Decide how to handle breaks and shift changes

A request made two minutes before lunch shouldn't count as a 32-minute response. Decide up front whether you count only working time. If you don't, the numbers will be skewed and people will stop trusting them.

Start with one area and one week

Pick the area where waiting hurts most, often material delivery or maintenance, and track it for one week, every shift.

On paper: a simple log at the point of request, with columns for time requested, who claimed it and when, and time completed. It works, but it depends on people filling it in during a busy shift, and the gaps tend to land on exactly the worst moments.

In a shared spreadsheet: better for analysis, but the same problem. Someone has to type the times, usually after the fact, from memory.

With a request system: the times are captured when people tap the request, claim it, and close it. Nobody has to remember to write anything down.

Look at the pattern, not the average

After a week, don't stop at one average. Look at three things.

Response time by shift. Second and third shift often look very different from first.

Response time by request type. Material, maintenance, quality, and supervisor calls usually behave very differently.

The ten longest requests. These usually point to a specific gap, like one responder covering too many areas, or a handoff at shift change.

Bring one number to your next improvement meeting

The goal isn't a perfect dashboard. It's replacing "it felt slow" with "maintenance requests on second shift waited this long before anyone claimed them." Once a number like that is on the table, the improvement conversation gets specific.

Tools that help

If you want a rough idea of what that waiting costs, the free downtime cost calculator at andontool.com/downtime-cost-calculator uses your own numbers, and nothing you enter is saved.

I built Andon Tool to capture these times automatically: requests from any phone or tablet, claimed by name, with response and completion times recorded and only working time counted, not breaks. There's a short intro video at andontool.com/demo, and if your plant is near Sioux Center, Iowa, I'm happy to come walk the floor with you.