Skip to main content

Stop Acquiring Users

While an experiment is running (active), you can choose to stop the experiment from acquiring any more users.

You may want to do this when:

  • You are satisfied with the experiment results but want to keep it running so the players can continue receiving variant configs. This means the events received from the game will continue to include A/B Test variant ids. These are available as Filters while querying metrics for the experiment duration

  • You want to start another (concurrent) A/B experiment and would like to direct new user traffic towards the new experiment, once you are satisfied the current experiment has acquired enough users

You can stop acquiring users from two places:

  1. Overview page actions

    A/B Overview

    • Select Stop Acquiring Users

    • Confirm you want to proceed

    Stop acquiring users

    • Status of the experiment will be updated

    A/B Status

  2. Results page

    • Select Stop Acquiring Users

    Stop acquiring users

    • Confirm you want to proceed

    • Experiment details section will be updated

When you stop acquiring users, the experiment will stop including any more new users. However, users who had already been assigned to a variant by then will continue to receive variant configs in their game. The metric results table will also be updated as long as the experiment is active. Please note that we require the experiment to run for enough time so it can acquire enough users in each variant for the statistical model to determine a winning variant (if there is one). If you choose to stop acquiring users sooner, this may result in the model not to run at all. For this reason, we recommend as a best practice that you keep the experiment running for at least a few days before you hit this action.

This is different from Stopping an experiment. When you stop an experiment, users that were part of the experiment will be removed and will no longer receive variant configs. Instead if there is an active remote config, they will receive the config values. The metric results table on the experiment details page will no longer be updated. Depending on whether by this time the model had calculated its results or not, the experiment will move to ‘completed’ or ‘canceled’ state.