# Lead time for changes: where the time actually goes

> DORA’s lead time for changes, split into stages: why opened-to-merged misses most of it, and what to measure to make it shorter.

10 November 2026 · Phillip Akuamoah-Boateng · 5 min read

Lead time for changes is one of DORA’s four key metrics. DORA defines it as the time a change takes to go from being committed to version control to running in production ([DORA’s four keys](https://dora.dev/guides/dora-metrics-four-keys/)). Teams with short lead times ship more often and recover faster.

Most teams can’t say where their lead time goes. The usual dashboards start the clock when a pull request opens, and that misses most of the story.

## Where a change spends its time

Split one change into stages:

1.  **Coding**: from the first commit until the pull request is ready for review.
2.  **Waiting for review**: from ready until someone first reviews it.
3.  **Review and rework**: from the first review until it’s merged.
4.  **Release**: from merge until it’s in production.

“Cycle time”, from opening to merge, only covers the middle two. A pull request that merges 12 minutes after opening looks great, even when its first commit was two days earlier. That hidden first stage is where big batches live: work piles up on a branch, and the pull request is opened when it’s all done.

**One change, stage by stage**

- Coding: 31h
- Waiting for review: 9h
- Review and rework: 5h

_Example change, for illustration._

## What makes lead time shorter

-   **Smaller batches.** Less coding before the pull request opens, and faster, more thorough reviews. See [why small pull requests get reviewed faster](https://nudgepr.dev/blog/small-pull-requests).
-   **A fast first review.** Waiting for review is pure queue time: nobody is working on the change. Give each pull request one person who’s accountable for reviewing it, and follow up when it waits.
-   **Short, predictable releases.** Automated deploys keep the last stage from undoing the rest.

## What Nudge measures

Nudge reads each pull request from your git host and records:

-   **Lead time, first commit to merge.** The clock starts at the pull request’s earliest commit. If its commits look newer than the pull request, because history was rewritten after it opened, the clock starts at the opening.
-   **Time to first review**, counted from when the pull request was ready for review. Time spent as a draft doesn’t count against the reviewer.
-   **Cycle time**, from opening to merge, so you can see how much of lead time happens before a pull request opens.
-   **Size**, and how review speed changes with it.

Insights shows the median of each, and compares your last 90 days before Nudge with the time since.

> **What Nudge doesn’t measure**
>
> Nudge doesn’t see your deploys, so it doesn’t measure the release stage. It measures lead time up to merge, which covers coding and review. If you release on every merge, that’s nearly all of it. Otherwise, add your release time from your deploy tool.

## Where to start

Compare your lead time with your cycle time. If lead time is much longer, work is piling up before pull requests open: work on smaller batches. If the wait for a first review is long, that’s queue time, and it’s the easiest stage to fix.

> **Try it on your own repositories**
>
> Connect GitHub and Slack, and see your lead time, review speed and pull request sizes for the last 90 days, before Nudge chases a single review.
>
> [Start free](https://app.nudgepr.dev/): 14 days, no card needed. The [quick start](https://nudgepr.dev/docs/quick-start) takes a few minutes.
