Skip to content
Open
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
28 changes: 19 additions & 9 deletions contents/docs/experiments/managing-lifecycle.mdx
Original file line number Diff line number Diff line change
Expand Up @@ -8,7 +8,7 @@ availability:
enterprise: full
---

After you've analyzed your experiment metrics and determined a winning variant, you can manage the experiment's lifecycle by ending, pausing, or resetting it.
After you've analyzed your experiment metrics and determined a winning variant, you can manage the experiment's lifecycle by ending, pausing, archiving, or resetting it.

## Ending an experiment

Expand All @@ -32,7 +32,6 @@ Remember, experimentation is an iterative process. Each experiment teaches you s

If you've already shipped a variant to 100% of users, you can still end the experiment using the **End experiment** button. In this case, the experiment is marked as complete without changing the feature flag.


## Pausing an experiment

You can temporarily pause an experiment to stop collecting data without ending it completely. This is useful when:
Expand All @@ -53,13 +52,24 @@ Technically speaking it resets the experiment's start and end dates. All events

This is useful when you need to restart data collection, for example after fixing a bug in your experiment setup or changing metric definitions.

## Comparing pausing, ending, and resetting
## Archiving and unarchiving experiments

Archiving removes an experiment from the default list view but preserves all data, including start/end dates and results. This is useful for keeping your experiment list clean after you've finished analyzing an experiment.

To archive an experiment, select **Archive experiment** from the **More** menu on the experiment detail page.

If you need to restore an archived experiment, select **Unarchive experiment** from the **More** menu on either the experiment detail page or the experiments list. This restores the experiment to the default list view without losing any data.

Unarchiving is different from [resetting](#resetting-analysis), which clears results and returns the experiment to a draft state.

## Comparing lifecycle actions

Here's how the actions differ:

| Status | Feature flag | Users see | Exposure tracking | Results | Can resume |
|--------|----------------------|----------|-----------|---------|---------|
| **Running** | Enabled | Multiple variants | Continues | Updating | N/A |
| **Pausing** | Disabled | Control variant | Stops | Fixed | Yes |
| **Ending** | Modified to roll out chosen variant to all users | Chosen variant | Continues (chosen variant only) | Fixed | No |
| **Resetting** | Unchanged | Multiple variants | Continues | Cleared | Yes (relaunch) |
| Status | Feature flag | Users see | Exposure tracking | Results | Can resume |
| ------------- | ------------------------------------------------ | --------------------- | ------------------------------- | --------- | --------------- |
| **Running** | Enabled | Multiple variants | Continues | Updating | N/A |
| **Pausing** | Disabled | Control variant | Stops | Fixed | Yes |
| **Ending** | Modified to roll out chosen variant to all users | Chosen variant | Continues (chosen variant only) | Fixed | No |
| **Resetting** | Unchanged | Multiple variants | Continues | Cleared | Yes (relaunch) |
| **Archiving** | Unchanged | Depends on flag state | Unchanged | Preserved | Yes (unarchive) |
Loading