This article explains how to set up an experiment. For more background about experiments, read About campaign drafts and experiments.
Things to note about experiments
- Experiments are only available for Search and Display Network campaigns. You won’t be able to create an experiment for Video, App, or Shopping campaigns.
- You can schedule up to 5 experiments for a campaign, but you can only run one experiment at a time.
- It may take some time for your experiment’s ads to complete the review process and to start running, depending on the size of your original campaign. You may want to schedule your experiment to begin in the future to prevent your experiment from beginning before your ads have been reviewed.
- You can decide what percentage of your original campaign’s budget that you want to allocate to your experiment. For Search campaigns in the new AdWords experience, you can also split your experiment by a cookie-based split or a search-based split. These experiments can help you see which campaign version performs better. You can then decide if you want people to see only your experiment or your original campaign, or show people both randomly.
- In Display campaigns, we always use cookie splits to ensure that users only ever see the experiment or your original campaign.
- As your experiment runs, you can monitor its performance.
If you make changes to your original campaign, those changes won’t be reflected on your experiment. Making changes to either your original campaign or experiment while your experiment is running may make it harder to interpret your results.
Keep in mind
If you select the cookie-based split option, and you use audience lists, you should have at least 10,000 users in the list to run an experiment. If you have less users than that, your results may be less accurate.
Before you begin
You need to create a draft before you set up a campaign experiment. If you haven’t yet created a draft, read Set up a campaign draft.
Set up an experiment
- Sign in to your Google Ads account.
- From the page menu on the left, click Drafts & experiments, then click Campaign Experiments at the top of the page.
- Click the plus button .
- Click Select draft, and select the draft you want to turn into an experiment.
- Enter a name for your experiment. Your experiment shouldn’t share the same name as your campaigns and other experiments.
- Choose a start date for your experiment.
- If you’d like to manually end your experiment, select None. Otherwise, choose an end date for your experiment.
- Enter the percentage of the original campaign’s budget that you’d like to allocate to your experiment.
- For Search campaigns, under “Advanced options,” choose how you want to split your experiment. You can decide whether you want to count users or searches when you split traffic between your campaign and experiment.
- Cookie-based means users may see only one version of your campaign, regardless of how many times they search. This can help ensure that other factors don’t impact your results, and may give you more accurate data.
- Search-based lets you can randomly assign users to either your experiment or original campaign every time a search occurs. You may get statistically significant results faster than a cookie-based split. It is possible that the same user could see both the experiment and your original campaign if they search multiple times.
- Click Save to finish creating the experiment.
- Once you create your experiment, you can find your experiments listed alongside your regular campaigns, as well as in your “Campaign experiments” page.
In some instances, you may encounter an error when trying to create an experiment. Here are a few reasons why this could happen, along with steps you can take to fix these errors.
- Your original campaign has been removed: Try creating a draft from a different campaign.
- Your original campaign already has a shared budget: Remove your campaign’s shared budget. Learn how.
- Your original campaign already has an experiment running or scheduled during your selected dates: Try selecting dates that don’t coincide with another experiment.
- Your campaign has features that aren’t supported by experiments: Try selecting a different campaign to convert into an experiment. Learn which features aren’t supported by experiments.
What you can do when you're ready to end your experiment
If you’re happy with experiment results, you can apply your experiment to the original campaign or convert your experiment into a new campaign.
- If you apply your experiment to your original campaign, the changes you were testing in your experiment will be applied to your original campaign.
- After converting your experiment into a new campaign, be sure to check the resulting campaign to ensure all changes were properly applied.
In both instances, your experiment’s performance data will be preserved. Learn more about how to apply your experiment.
If you prefer to simply end your experiment, you can do that by changing the end date of your experiment so it stops running at the close of the day. You can also click Edit in the upper-right corner of the page and select End now. Learn more about making changes to your campaign experiments.