Testing is succeeding.

What is an A/B test?

An A/B test is a type of experimental marketing research. It involves two equal but different stimulus (given by some special conditions) which have different outcomes. The idea is to understand if one of the two stimuli performs better than the other considering only one change in one of the elements as a variation.

For example, you have to ads with exact same visual imagery and text, and the only thing that you change is the call to action of one of them. A/B testing will give you the insight to understand if one call to action is better than the other, or if it works equally for both.

How does it works?

Considering two exact same pieces, you only vary one element between the creative execution of a specific touchpoint that can be done in websites, ads, e-mails, direct mail, landing pages, ads, SMS, etc.

All of these pieces are constructed by several elements, in an A/B test you should choose only one element to test. You can vary only one element and understand how your audience reacts to different exposure of this specific element. Variations can be of image, copy, call to action. target, headline, layout, subject, etc.


The clue is to only vary one element, the rest of the piece should remain the exact way, so you really can attribute the results to that specific variation.

Decide the variable you want to test, choose the audience that you are going to use to make the experiment, set the time the test will last, check that there is only one variable that changes from piece A to piece B. Run the test. Analyze.

Do all the testing you need.


  • It is not as expensive as it seems and it is very common in online ads.
  • Testing gives you great insights to understand if you are using creative elements to achieve your goals.
  • You can test multiple times, to really understand your audience and change your content to be successful.
  • Results are quick and is easy to analyze.

It seems pretty easy to do, but you need to have your intention very clear. You need to understand what do you want to test, choose the element you will change, and understand to which audience it will be shown. Analyzing results it is very important to give the test a sense and take value.

Hope you liked today’s “Let’s talk about…” topic, I look forward to reading your comments and positions about it! If you want to know more about marketing insights, check this post!

You Might Also Like

Leave a Reply