Ending an Experiment

When an experiment ends, one of two situations typically applies:

• Your experiment has statistically significant results and it’s time to ship an experience (control or test) to all your users.
• Your experiment has a problem and you need to stop it.

Making a decision

Stopping an experiment

If your experiment has a problem and you need to stop it, there are two options.

• If the problem means you no longer want to run this test, Abandon the experiment. This puts the experiment in a finished state and all users fall back to defaults in code. You can still Reset it later if you decide to revisit the experiment.
• If one of the groups has a bug you plan to fix and you want to restart the experiment later, Reset the experiment. Alternatively, you can disable a group with the bug if you want to keep the other test groups running.

Resetting puts the experiment into an unstarted state. Every user receives the default experience (either the value defined in code or the layer default). The "salt" used to randomize a user’s group also changes. When you start the experiment again, Statsig randomly assigns users to a group that isn’t necessarily the same group they were in before the reset. This ensures that a group that didn’t perform well due to a bug or bad experience in the previous run isn’t negatively affected after you address the issue.

Archiving an experiment

After you make a decision on an experiment, or when you abandon or reset it, you can archive the experiment. Archiving preserves the experiment’s history and results and makes the experiment read-only.

If you later decide to make the experiment active again, unarchiving the experiment makes it editable but doesn’t automatically restart it. If the unarchived experiment was previously part of a layer and you intend to include it again, you must manually reset allocation from that layer.