Custom experiments let you propose and test changes to your search or display campaign. Experiment sync ensures that any optimisations you make to your original campaigns while your experiment is running, will be automatically copied to the experiment (trial) campaigns.
This article explains how experiment sync works and how to set it up.
How it works
This manual process of copying changes to experiments can be time-consuming and error-prone, especially if you are running multiple experiments at the same time.
When your trial starts at the scheduled time, experiment sync will:
- Automatically update the trial with any changes made to the base campaign since you created the trial.
- Changes on the base campaign will be automatically synced to the trial campaign. However, changes resulting from this sync aren’t currently reflected in the 'Change history'.
Note: Experiment sync only works when you make changes in one direction, from base campaign to trial campaign. If you make changes in your trial campaign, those will be ignored.
The trial campaign is a separate snapshot of the base campaign at the time it was created. Changes made in the base campaign will be independently applied to the trial campaign. The trial campaign will reflect the exact changes you make, therefore it may not mirror the overall state of the base campaign.
For example, if you remove all languages from the base campaign, it will result in targeting all languages by default. The trial campaign won't necessarily switch to targeting all languages by default. Instead, it will only remove the languages you excluded during sync.
| Date | Changes made by advertiser | Status of Sync | Base campaign | Trial campaign |
|---|---|---|---|---|
|
2 November |
Created an experiment with sync enables | Enabled | Targets languages A and B |
Targets languages A and B |
| 3 November | Added languages C and D in trial campaign | Enabled | Targets languages A and B | Targets languages A, B, C and D |
| 7 November | Removed languages A and B from base campaign | Enabled | Targets no languages | Targets languages C and D |
Bear in mind
Experiment sync is only applicable to custom experiments in search and display campaigns. Any changes made to the trial campaign won’t be synced with the base campaign. Experiment sync is turned on by default when you create a new trial (you will have the option to turn sync off while creating the trial, but can’t change the setting after the trial campaign has been created).
After experiment sync is turned on:
- It can’t be turned on for trials that have been scheduled or are already running, but it will be available and on by default for new trials.
- It might require up to 10 minutes after the trial has started to implement the changes.
- If you have more than one trial scheduled to run (for example three trials starting at different times), experiment sync will update all trials with the same updates to ensure that they are aligned with the base campaign.
What happens if an error occurs
If a sync error occurs during the trial, the system will stop syncing changes for the remainder of the trial. You’ll need to manually sync the base campaign changes to the trial. For example, if you've changed the name of an ad group in the base campaign, and the trial campaign already had an ad group with the same name.
Instructions
- Follow the instructions to set up a custom experiment.
- In the 'Advanced options', open the 'Experiment sync' section. It will be set to On by default.
- Click Create experiment. Your trial is completed and ready to run.
