What is the difference between cycle time, lead time and touch time?

Why do these three get confused?

Because in everyday speech they all mean "how long it takes". On a depot floor they are three different clocks, and an improvement plan that uses the wrong one will improve the wrong thing.

I was caught by this myself in a practice set earlier this year. The question asked which measure changes when the constraint is improved, and I answered with the wrong one. The habit of keeping the three apart is something I now practise deliberately.

What is cycle time?

Cycle time is the time one unit takes from the start to the finish of one step. If a receiving clerk takes a pallet from the dock door, checks it against the booking and scans it in, the cycle time of receiving is the time from picking up that pallet to putting the scanner down.

It belongs to the step. It says nothing about how long the pallet waited at the dock door before the clerk reached it.

How do you measure it?

With a stopwatch, on real units, several times, across more than one shift. One timing is an anecdote. A spread of timings tells you how stable the step is, and an unstable step is often more trouble than a slow one.

What is lead time?

Lead time is from request to delivery across the whole flow. For an inbound delivery it starts when the vehicle is booked in and ends when the stock is available to pick. It is the measure the customer feels, because the customer does not see your steps, only the wait.

Lead time includes every queue. That is the point of it.

Why is lead time usually much longer than the sum of the cycle times?

Because of waiting. Units sit at the dock door, sit in the put-away lane, sit in staging. Add up the cycle times of the steps and you get a number that looks respectable. Measure the lead time and you get a number several times larger. The difference is not a mistake. It is the queue.

What is touch time?

Touch time is the hands-on work inside the flow: the minutes in which someone is actually doing something to the unit. It is usually a small fraction of the lead time, and showing a client that fraction for the first time is one of the more useful moments in a review.

Where does waiting hide?

In the gap between lead time and touch time. If a delivery has a lead time of a shift and a touch time of a few minutes, almost all of that shift is waiting. Waiting is what the customer feels, and it is also the cheapest thing to remove, because nobody has to work faster. The work simply has to stop sitting still.

Which measure changes when you improve the constraint?

Throughput: the rate at which finished work leaves the system. Not the cycle time of the fastest step, not head count, not the number of procedures. If a change at the constraint does not move throughput, it was not a change at the constraint.

This matters for improvement plans. A plan names a measure, and if the named measure is the wrong clock, the review at the end will show effort and no result.

How do I use the three together?

• Map the flow and mark the queues.
• Time the cycle time of each step on real units.
• Measure the lead time end to end.
• Work out the touch time and compare it with the lead time.
• Find the step where work queues longest. That is where to look for the constraint.

Q&A

The questions I am asked most about these three measures are answered below.

Last updated 1 October 2026.