The Don’ts in A/B Testing

The Don’ts in A/B Testing

Lemuel Galpo
Written by Lemuel Galpo
October 10, 2014
Do and Don't in A/B testing

In today’s virtual world, there is no denying that a lot of entrepreneurs need more than just their old campaigns and techniques. Since there are a lot of changes made in the web on a daily basis, the good old days when optimizing search engines was one of the most effective ways to attract global online users has now come to an end. Aside from SEO, A/B testing is another new tool that web owners and developers can really use to their advantage. As Josh Frafchin discussed in his webinar, there are a lot of do’s and don’ts when it comes to A/B testing

Now, one thing that you must remember as we go further, is that the number one aim of A/B testing is to make the webpage profitable. I am reminding you of this because so many people get lost in the process of A/B testing. that they forget the main objective is to make the page as lucrative as possible. Getting a tight grip on its main purpose would enable users to work towards maximizing the positive effects of A/B testing.

A/B Testing Defined

Okay, let’s go back to the days when you needed to complete an experiment for your science project. If you can still remember, there are some variables that need to be tested during a science project. Through the experiment, you would be able to know which variable works best. This is exactly just like A/B testing. From the term itself, A/B testing enables you to test elements A and B. Through running your tests, you would be able to see which variable is more feasible for your page.

To further simplify what A/B testing really is, it is one way of helping you figure out whether the changes that you would apply in your current webpage would be advantageous or not. With the help of A/B testing, coming up with a hypothesis and making wise and well-rounded decisions will not be a difficult task anymore. Even if you are still a novice at heart, A/B testing can help you make intelligent decisions along the way, making your page more profitable.

Where to Use A/B Testing

What you need to test in your webpage highly depends on what you want your clients to do. For instance, if you want to make people click that ‘buy now’ button, you have to explore the different elements in your page that might be stopping them from doing so. Through this, you would be able to determine the weakest points in your page and work towards its improvement.

The elements that are frequently tested include the length and number of the texts in a single page, costs and promos, call to action, photos of products, length of the form and field type and lastly, product title and description. By exploring these areas, you can identify what your audience really wants to see on your page.


What Not to Do During Testing

Do and Don't in A/B testingAs mentioned previously, Josh Krafchin has outlined what to do and what not to do in A/B testing. The results of these guidelines can also be seen on his website; Clever Zebo. Apart from knowing what to do and how to use A/B testing, it is also important to know what not to do to avoid any mistakes or errors. Here is a list of what not do during A/B testing:

  • Do not Rush. Test One Element at a Time: If you are going to run A/B testing, you have to run tests on both sides at the same time. One common mistake of some people is that they test element A first before element B, and this yields to inaccurate results.
  • Don’t Shock Your Web’s Patrons: If you have to run tests on a very big part of your webpage, it is best that you exclude old visitors of your site in your test so as not to confuse them. Therefore, including new visitors is a better option especially if you are about to test on the core elements of your page.
  • Don’t Always Stick to Your Intuition: The results of A/B testing may be very surprising at times. If you feel that the results do not match with your intuition, stick to the results because that’s the main purpose of the experiment.
  • Do not be impatient. Patience is Always a Virtue: When running A/B tests, you should always be patient and wait for the test to be completed up to the end. Some people tend to conclude early on during the experiment. This beats the purpose of running the test because you would not be able to see accurate results from the test.

[Tweet “A/B testing lets you see the big picture in a clearer perspective”]

If you want to make adjustments in your page, A/B testing is an option before making any decisions. If you want to learn more about A/B testing and the like, you can check out the entire webinar here.

Originally published October 10, 2014 - Updated April 15, 2019
Lemuel Galpo
As Customer Content Manager, Lem is responsible for bringing learnings in conversion optimization and testing to the world. He is part of's growth team and coordinates all writers, editors and illustrators.
Guest Post Form

We have brought thought leaders, influencers, visionaries and veterans to our tribe. Now it’s your turn. If you have something worthwhile to share with a large community of savvy testers, go ahead and pitch your post idea. We’re listening.



[hclightbox id='5' text='Anchor text']