Increase AdSense Earnings by changing ad settings with Experiments

Create and run A/B experiments for any live AdSense Ad unit of your website/blog and analyze which ad setting is working for you to increase earnings.

It is a well-known realistic fact that a particular AdSense ad settings works for one website/blog that might not fit for another site. This is because AdSense earnings depends on lots of factors including your site’s layout, contents, traffics, number of Advertisements, placements of Ad Units and its settings etc. Hence, no one can suggest best practice for your AdSense ad unit that can boost your income.

Therefore, publishers had to change settings of specific ad unit for A/B experiments and place the changed AdSense ad setting codes on their site for some unknown period say 1 day, 7 days or even a month or months. Thereafter, they will be able to compare and choose best performance (i.e., the highest earning) of any one of both settings of different periods. This process results many confusions to select one of those settings that will best fit for their website.

The worst thing of these types of testing is that you can’t compare the performances of both changes within same period with your live AdSense Ad Unit. Additionally, most of your classical A/B testing will lower your earnings.

But, now you can run an Ad unit on your site with two different settings at the same time and will also be able to identify best settings i.e. winner of your experiment for that particular ad unit after getting reliable confidence. This feature is known as AdSense experiments. In other words, you may now run scientifically A/B researches with a new AdSense experiments feature available in your AdSense account.

Before we use experiment tool available in AdSense account let us decide exactly what we would like to test in the experiment. There are two types of experiments viz. Allow & block Experiments and Ad unit Experiments. Remind that you can’t run both types of experiments simultaneously.

A) Allow & block Experiments will let you decide different variances under General category, Sensitive category and which type of ads you are willing to show e.g. user-based ads, display ads and text ads etc.

B) Ad unit experiments will let you choose different ad types and text ad style settings for your A/B split testing.

I have created ad unit experiment and run it for 57 days to achieve 92% confidence and choose variation as a winner which boost my AdSense earning by 79 percent. Therefore, I am sharing my experience with all of my readers so that it may help them to increase their AdSense earnings.

How to Use AdSense Experiments Feature to Increase your Earnings

This process will help you to use AdSense experiments tool to compare one of your ad settings against a variation of that setting to see which performs better. To create a new experiment in your Google AdSense account, follow these simple steps:

Step-1: Go to My ads tab of your AdSense account and under “contents” menu in the left navigation select Experiments.

Step-2: You will see two segments to create your research there, one is “Allow & block Experiments” and other is “Ad unit experiments”. You have to click on Create button under any of these two segments depending on which type of experiment you want to create.

Step-3: Give any name to your experiment. The goal of your experiment is obvious i.e. to find out whether your original ad setting or the variation will maximize revenue.

Step-4: Enter a brief description of your experiment in the Note box for future reference. This is optional and therefore you may left it blank.

Step-5: If you have decided to create Allow & block Experiments then follow step-5A and if you have chosen the second section i.e. Ad unit experiments then follow steps-5B

Step-5A: Create Allow & block Experiments

In this type of experiment you may allow or block ads from different products under three types of variation sections viz. general category, sensitive category and ad serving settings.

Click Add to experiment button in the appropriate section and in the open window click on Add button next to each product to add it for your experiment.

After adding products from the category or ad serving opt-out that you want to compare in your experiment you will see the variation column automatically fills with blocked or allowed depending on the original product settings.

Step-5B: Create Ad unit experiments

In this type of experiment you may make variances with all ad type and text ad style settings. From the Choose an ad unit drop-down in the Ad unit section select any ad unit that you want to add in the experiment.

However, following ad units are not eligible for use in experiments:
i) Idle ad units. Idle ad unit means the ad unit that received no impressions within the last 7 days.
ii) Ad units that are already being used in another experiment.

In the variation section, select the Ad Type in accordance with your ad unit setting you want to compare in your experiment. For example, if your original ad unit is Display ads only, you could set your variation to be Text & display ads or Text ads only.

Thereafter, you may also change Text Ad Style for that particular ad unit. You may either select any one ad style from the different styles given there or you may use custom settings to create a new ad style.

Step-6: I already told all of you that with the help of this AdSense experiments tool you will be able to use two different ad settings at the same time on your live Ad Units. Accordingly, your site’s traffic should be divided between your original ad setting and the variation.

Consequently, in the Traffic split section you may choose how often you would like to show the variation. The default variation is set for 50% of the time. You may change it to show variation from 10-90 percent.

But, I personally recommend all of you to let Google optimize the traffic split. This is because Google uses multi-armed bandit to adjust the amount of traffic on daily basis depending on the real performance differences of the original and the variation and does not take just a random chance.

Step-7: Finally, click on Start experiments to move your experiment into running stage on your live AdSense Ads.

How to see the performance result of AdSense Experiments

You will see the initial results of your experiment next day. Therefore, after 24 hours visit My ads and select Experiments then click on the name of your experiment to see the performance result of AdSense Experiments. You can also access your testing report from the Performance reports tab in your account. See the following Screenshot of performance report of running AdSense Experiments after 3 days.

How to Choose the Winner of AdSense Experiments

You may choose the winner at any time to see the confidence level. However, it is recommended that if one of the settings has reached at least 95% confidence (indicated by five blue dots), you should choose the winner of your experiment.

Once you have chosen the winner, your experiment moves from the running state to the completed state. You may also stop/end your experiment at any time. Please fill free to share your own AdSense experiments with us.