Lead time vs. cycle time: how to measure both

01

Set boundaries for each clock

For this guide, lead time runs from a customer's request to delivery. Cycle time runs from the start of work on that request to delivery. Both are elapsed times and can include waiting. If your team uses different definitions, write them on the report and keep them consistent.

Capture the request, work-start and delivery timestamps for each item. Specify whether the clock uses calendar time or business time and how cancellations and items still open are reported. Measure hands-on effort separately if you need to understand labor use.

02

Calculate and compare

Hypothetical example: a request arrives Monday at 9 a.m., work starts Wednesday at 9 a.m. and delivery occurs Friday at 9 a.m. Using elapsed calendar time, lead time is four days and cycle time is two days. The two days before work started are part of lead time but outside the defined cycle time.

The cycle time can still contain internal waits; it is not automatically two days of active labor. Examine individual records and show a distribution or age bands as well as a typical value. Report open items separately so slow, unfinished work is not hidden.

03

Use the difference to investigate

If the gap between the clocks grows, inspect intake and the queue before work starts. If cycle time grows, trace internal handoffs, interruptions and returns. Compare similar work types and periods because a more complex mix can change both measures.

Pick one suspected cause, agree a limited change and monitor delivery time alongside quality and completion counts. Check whether the change shifts waiting to another step or excludes difficult work from the measurement.

For a useful comparison, group requests by type and choose a completion cohort, such as items delivered in a given month. Compare medians and the longest waits within that group, and disclose the number of items. A completed-item view can look faster while an old backlog remains open, so also report the number and age of unfinished requests.

Decide how to treat an item returned after delivery: record the original delivery and the later correction separately, linked to the same request. If customers care about the time until an acceptable result, add a clearly defined measure for that outcome. Do not silently replace the original delivery timestamp, since doing so obscures what happened.

04

Build a timestamp worksheet

Create one row per request with request time, work-start time, delivery time, work type and status. For completed work, calculate both elapsed intervals from those timestamps. For open work, show the current age separately rather than using today as if it were a delivery date. Check that time zones and business-hour calendars are consistent.