Use A/B or split testing for Facebook Ads

If you have been working in a function such as web development or marketing, chances are that you have used A/B testing, also known as split testing, in one of your projects. A/B testing is a methodology, especially popular in advertising, that uses randomized experiments with two variants, A and B, in a controlled experiment.

Like most other online advertising, Facebook ads can be difficult to master. But with the help of A/B testing, you can soon start getting insights that would help you create better campaigns, and next time you want to run an ad on Facebook, you will spend lesser and lesser time creating successful ads.

In this post, I will walk you through all the nitty-gritties you need to know about conducting an A/B testing with your Facebook ad –

Create a campaign –

The first step is, of course, creating an ad. But the important thing to note here is that when you want to split test a few ads with a particular theme, bucket them all under one campaign name. This way, it is easier to track the performance of the ads and it is also easier to organize the ads. Doing this is easy in Facebook. All you have to do after creating one ad is to open it and click “Create a Similar Ad”, as shown below –

Once you do that, any number of similar ads that you create will all show under one campaign name as below –

Change one variable at a time –

This one is important and you should always try to change only one variable at a time, i.e. test your ads for only one element that you want to test for. Keep all the other elements same in the A/B tests. In the example below, I wanted to test the copywriting. All the other details of the ads were the same. Believe it or not, the ad on the top had 53 clicks and the one in the bottom had 0 clicks.

Try different landing pages – 

One element you can test, as mentioned above is copywriting. Another element that you can test in your A/B test is the landing page associated with your ad. For example, in the ad below, the one on the top had an external website as the landing page, and the ad below had the Facebook page as the landing page. Although first ad got a few clicks, none of them translated into a page like post like, or any other activity that could be useful.

Run the A/B test ads during the same time –

There is a direct correlation between the clicks, activity and the time that you run your ads. Thus, you should always run the test ads during the same time. Also, test for the same time duration. Otherwise, the results would be meaningless.

Monitor the results closely –

Facebook has built some great insight tools. Take all the advantage you can of these tools. When you are running these ads, make sure that you have enough time at hand to monitor the results closely. Say, you decided to start the ad right now. decide a time in advance when you want to come back and see the results. If you decided to check the analytic after 1 hour, stick by that deadline and analyze the results. This would help you to improvise fast and get better results soon.

Test your target audience too –

When I create ads, I like to play around with the “targeting” option too. Facebook lets you give options for the interests that you want to target. Say your audience is sales and marketing professional, in one ad try “#sales” and in the other “#marketing”. Even though your product is meant to appeal both kinds of audiences, may be your copywriting is such that more sales people click on your ad or vice versa. You would be pleasantly surprised by the results when you try these for yourself.

You can always try testing with other factors such as age, gender, etc. when you create ads, but above are the key elements that you should definitely keep in mind when you think about A/B testing the next time.