What it means
Every process, in a factory or an office, has a cycle time: the elapsed time between a unit entering and the same unit leaving. The term is used with two related meanings, so it is worth saying which one is meant.
In manufacturing engineering, cycle time is often the time a single workstation takes to process one unit, and the process cycle time is set by the slowest station, the bottleneck; in process improvement and service operations it usually means the total elapsed time for a unit to pass through the whole process, which is sometimes also called lead time or flow time. Cycle time is made up of touch time, the time work is actually being done on the unit, and everything else: waiting in queues, waiting for information or approvals, being moved, being inspected, being reworked.
In most processes the touch time is a small fraction of the cycle time; an insurance claim that takes three weeks to settle might receive two hours of actual work, and a purchase order that takes ten days to approve might take twenty minutes to process. The ratio of touch time to cycle time, sometimes called flow efficiency, is often below 10%, and this gap is where improvement lies: the work itself does not need to be faster, the waiting needs to be eliminated.
Three measures work together: throughput is the number of units completed per period, and in a stable process equals one divided by the bottleneck cycle time, while work in progress is the number of units in the process at any time. Little's Law connects them: work in progress equals throughput multiplied by cycle time, which is exact for any stable process and means that for a given throughput, cycle time can only be reduced by reducing work in progress, because piling more work into a process does not make it faster, it makes it slower.
Takt time, a fourth measure, is the rate at which customers demand units (available time divided by demand), and it sets the target that the bottleneck's cycle time must meet. Reducing cycle time delivers several kinds of value.
Shorter cycle time means faster response to customers, which supports higher prices, better retention and the ability to promise delivery dates competitors cannot; it also means less work in progress, which in manufacturing is inventory that ties up cash and in services is a backlog that generates chasing calls and errors. It means problems are detected sooner, because the gap between a defect being made and being found is shorter, and capacity is released: the same people and equipment complete more units, so growth can be absorbed without proportional cost.
The tools for reducing cycle time are those of lean and process improvement: mapping the value stream to see where time goes, identifying the bottleneck and protecting it from starvation and disruption, balancing the workload across stations so that no station waits for another, reducing batch sizes so units do not wait for their batch to complete, eliminating handoffs and approvals that add no value, and standardising work to reduce variation. The finance function's role is to measure the cycle time of its own processes (order to cash, purchase to pay, month-end close) and to help the business value the improvements elsewhere.
In practice
Real-world examples
Example 1
An accounts payable team measures the cycle time from invoice receipt to approval at nine days, with twenty minutes of touch time, and reduces it to two days by replacing sequential email approvals with a single workflow step.
Example 2
A software team tracks cycle time from starting a feature to releasing it, finds a median of 23 days with most of the time spent waiting for code review, and halves it by limiting work in progress so reviews happen immediately.
Example 3
A hospital measures the cycle time from patient arrival to discharge in its emergency department, identifies waiting for blood test results as the largest component, and installs point-of-care testing.
Think of it
Cycle time is how long it takes to complete something from beginning to end-your process speed.
Formula
Calculation
Cycle time (per unit) = Elapsed time from start to finish of one unit
Process cycle time (rate) = Cycle time of the bottleneck station
Throughput = 1 / Bottleneck cycle time
Little's Law: Work in progress = Throughput x Cycle time, so Cycle time = Work in progress / Throughput
Takt time = Available production time / Customer demand in that time
Flow efficiency = Touch time / Total cycle time x 100
Worked example. A furniture workshop makes a chair through five stations with these times per unit: cutting 6 minutes, assembly 12 minutes, sanding 8 minutes, finishing 15 minutes, packing 5 minutes. Total touch time is 46 minutes.
- Bottleneck: finishing at 15 minutes; process capacity = 60 / 15 = 4 chairs an hour = 32 chairs in an 8-hour shift
- Demand is 40 chairs a day; takt time = 480 minutes / 40 = 12 minutes; finishing at 15 minutes cannot meet it, so the workshop falls 8 chairs a day behind or works overtime
- Work in progress on the floor is 120 chairs; by Little's Law, cycle time through the workshop = 120 / 32 = 3.75 days, against a touch time of 46 minutes; flow efficiency = 46 / (3.75 x 480) = 46 / 1,800 = 2.6%
Improvement. A second finishing booth halves finishing's effective time to 7.5 minutes. The new bottleneck is assembly at 12 minutes, exactly at takt: capacity 40 chairs a day. Work in progress is then capped at 60 chairs through a pull system; cycle time = 60 / 40 = 1.5 days.
Financial value. Eight additional chairs a day at $250 contribution each = $2,000 a day, about $500,000 a year over 250 working days, against a booth costing $120,000 and an operator costing $45,000 a year. Reducing work in progress from 120 to 60 chairs releases 60 x $180 of materials and labour = $10,800 of cash, and halves the space, damage and expediting associated with the backlog. The revenue effect dwarfs the inventory effect, which is typical: cycle time improvements pay mainly through capacity and responsiveness.Case study
Seen in the real world
This is a fictional, illustrative scenario. An insurance company's motor claims department settled 40,000 claims a year with an average cycle time of 21 days from notification to payment. Customers complained, renewal rates among claimants were poor, and the department's cost per claim of $85 was above the industry average. A value stream mapping exercise followed fifty claims through the process and found that the average claim received about two and a half hours of actual work: the remaining twenty days and several hours were spent waiting, in eleven separate queues, for allocation, for documents, for an engineer's report, for a supervisor's authorisation, for payment processing. Eighteen percent of claims were reworked because information gathered at the first call was incomplete and had to be chased.
The redesign attacked the queues rather than the work. A structured first call captured all the information needed, with photographs uploaded during the call, which cut rework to 5%. Authority limits were raised so that handlers could settle claims under $5,000 without supervisor sign-off, removing a queue that had averaged four days. Engineer reports were replaced by photograph-based assessment for minor damage. Work in progress was capped per handler, so new claims were not started until existing ones were moved on, and a daily review dealt with any claim older than ten days.
Cycle time fell to six days within four months, and the department's cost per claim fell to $62 as chasing calls, rework and supervisor time disappeared: a saving of $23 x 40,000 = $920,000 a year. Customer satisfaction after a claim rose sharply and renewal rates among claimants recovered to the level of non-claimants. The head of claims noted that the department had not hired anyone, bought any system or asked anyone to work faster; it had simply stopped making claims wait.
Watch out
Common mistakes
- Trying to reduce cycle time by making people work faster, when touch time is a small fraction of the total and the waiting between steps is where the time goes.
- Starting more work to "keep people busy", which raises work in progress and, by Little's Law, lengthens cycle time for everything in the process.
- Improving a station that is not the bottleneck, which changes nothing about throughput and may simply build a queue in front of the bottleneck.
Questions
People also ask
What is the difference between cycle time and lead time?
Usage varies. Commonly, lead time is the elapsed time from the customer's order to delivery, which includes time before work starts; cycle time is the elapsed time from starting work to finishing it. In manufacturing engineering, cycle time can also mean the time per unit at a single station. State the definition in use.
What is the difference between cycle time and takt time?
Cycle time is how long the process takes per unit; takt time is how long it may take per unit to keep up with demand. A process whose cycle time exceeds takt time falls behind; one well below takt has spare capacity.
How does cycle time affect cash?
Through work in progress. In manufacturing, work in progress is inventory that has consumed cash for materials and labour but has not yet been sold; halving cycle time at constant throughput halves it. In services, the effect is on receivables: a faster order-to-cash cycle brings cash in sooner.