How to Setup Experiments in Google AdSense?

AdSense is a popular pay per click program from Google that allows publisher to monetize their online content. Google offers large variety of ad formats to place the ads in different places on a website. Availability different ad formats make it difficult for publishers to choose the high revenue generating ads. In order to solve this problem, Google offers experiments option to do A/B testing with ads on your live traffic.

What is AdSense Experiment?

AdSense experiments are A/B testing between two ads to find the best performing one for your site. You can create a new ad variation and run it along with the original ad to find the better one. For example, you can test two different ad formats on the same place on your site to check which is performing well. This will help you to get high revenue by optimizing ad display. You can carry out different types of testing with experiments option:

• Run experiment on ad balance to get optimized ad fill rate
• Test with different categories included or excluded
• Change ad styles like color and background for testing
• Choose text, image or both and check which one performs better.

If you are an AdSense publisher for long time then this a definite opportunity to try out.

How to Setup Experiments in Google AdSense?

Login to your Google AdSense account. If you have an opportunities to improve revenue, Google will automatically shows them under “Optimization > Opportunities” section. Otherwise you can go to “Optimization > Experiments” and start running your own A/B testing.

The experiments section contains three parts:

• Auto experiments
• Custom experiments
• View status and results

1. Enabling Automatic Experiments by Google

You can enable the slider allowing Google to run automatic experiments on your site. Google will run testing on 10,000 daily ad impressions. If your site gets lower impression than 10K on daily basis then Google will test on all the impressions, which may not be good for testing. So, we recommend you to try this if your site gets good traffic.

Google will run the automated experiments on 50% of the impressions and run the original ads on remaining 50% impressions. You can compare the results to find which option is better for you to generate more revenue.

2. Creating Your Custom Experiments

If you don’t want to Google to run automated experiments or you like to check custom scenarios along with automated run then you have few options.

2.1. Ad Unit Settings

Ad unit settings will allow you to choose any of your ads and run test by changing the type or style. Click on the option and fill out the required details to create an ad unit experiment.

• Experiment name – enter name for your experiment, for example – Ad unit xxxx test1.
• Ad unit – click on the dropdown and choose one of your custom ad that you have created and placed the ad code on your site.
• Experiment settings – this is an important selection. You can select ad unit type or text ad style for testing. Google will list down further options based on your selection.
• Basically the style will show you the options for selecting one of the default ad styles or option to create custom ad style. You can customize the new style and run experiment with the original style of the ad.

• Ad unit type will allow you to create variations for text, image or text/image ad types.

• You can enable to choose the winner automatically or choose the winner after Google completes the experiment.

2.2. Blocking Controls

This is a second level of setting up experiments based on ad categories. The process of filling the details remain same like above with the difference that you need include or exclude categories for your variation. For example, if you have sensitive category excluded in original ads then enable in the variation and run the experiment. The result will show you whether including sensitive category ads can help in improving the revenue on your site.

2.3. Ad Balance

Ad balance is a tricky feature that helps primarily on larger sites. This will show you the revenue versus ad fill rate and recommend the possibility of reducing the ad serving on your site for better user experience. Instead of enabling this straight away, try to run an experiment to test whether it helps you.

Google allows creating only one ad balance experiment at a time since this runs at account level. Click on the “Ad balance” button and enter the name for your experiment. Under “Experiment settings” section, adjust the “Ad fill rate” slider as recommended.

Enable or disable, “Automatically choose winner” option and click on “Create” button. The experiment will start immediately for the duration decided by Google. You can monitor the status and choose or cancel the experiment winner or original ad at any point of time.

3. View Your Experiments

You can view all the experiments under your account in this section. Here you can get the summary of status, dates and progress for each experiment. If you want more details about an experiment then click on the name and get the result.

We recommend not choosing the winner automatically, so that you can analyze the result and decide to select the original or variation. If required you can continue the experiment for more time before deciding.

Conclusion

Now you have learned how to create and setup experiments for ad level, category level and at account level on your AdSense account. Always try to setup the experiment for various styles, type, categories and balance to choose the best revenue generating ads for you. If you don’t want to have any of these hassle simply disable all your custom ad codes on your site and enable auto ads. Google will consider all these factors automatically and serve the ads on your sites without any actions needed from your side.