4 Ways To AB Test Your Display Ads Right Now

Want to know how to AB test your banner ads? We’ve got the answers.

Online advertising has become much more demanding, meaning any changes you can make to your ad banners can, potentially, have a huge impact on your overall campaign performance. But how do you know what changes you should make? This is where AB testing comes into play.

Rather than relying on your gut feeling, your opinion, or your friend’s opinion (as valuable as all these things may be) really, the most important thing is what actually works.

While, yes, there are universal truths and guidelines for creating good ad banners, you should always be led by what the data tells you. So, what should you AB test then?

AB Test for Max Impact & Increased ROI


Experiment with different background colors and formats (i.e, light, dark, strong, subtle) Also, think about arranging elements differently on your visual display. Ask yourself questions like, “does the use of photos vs. illustrations change audience engagement ratios”

Ad Copy

Experiment with different wording in your ad copy. Would your ad perform better by emphasizing the benefits of your product or service? Or would it work better by emphasizing its features? Also, don’t forget you can play around with the formatting of the text like different fonts/sizes, etc. 

Calls To Action (CTA)

As covered in our last post about creating awesome ad banners, your creative should have a clear CTA. But you can test different wording, like “Join Now” or “Sign Up”. Does red really work the best for a CTA? On that note, referrals, reviews, and testimonials from customers who have already tried and tested that product are dynamite selling points when added to the bottom of a banner ad. Perhaps try one banner with a testimonial and one without. Compare the results of each and stick with the one that boasts that highest engagement moving forward.


Using different pricing levels or promotional offers is a good thing to test in your ad creatives. For example, using $0.99 rather than $1.  Test which one will drive more engagement and ask yourself if offering free shipping or a discount on the final basket total might work better. 

Bes selective when testing ad elements

Don’t modify and test too many different elements at once. If, for example, you want to test ad copy, then stick with this only and avoid jumping into other factors like design elements, pricing, etc. Reason being that if you change too many things at once, it’s going to be much harder for you to see later what changes led to improved results.

Give your Split Testing time to track

AB testing can sometimes seem like a daunting task but it’s a vital step for creating ad campaigns that convert well.  However, think of it this way. Rather than breaking your campaign down into its core concept, creation, targeting, and optimization phases, think of the “Optimization” phase as an overall guiding principle that dominates and directs the overall flow of the campaign.

You should also avoid jumping to conclusions too early. Depending on the volume of your traffic, you should let your AB test run for around 4 weeks, continuously, then look at the data to understand which one performed better.

Looking for top tracking services in the market today to help you split test, rank, and optimize your online advertising traffic? Two of our favorites are Voluum and RedTrack.

Start Split Testing and get the most out of your online advertising

It may seem obvious to test and optimize every element of your ad campaigns, but a lot of companies fail to do so, meaning they’re missing out on one of the huge advantages that digital advertising has to offer. Tha bilitity to gather information and insights at the click of a button and the oppurtunity totest and improve constantly.

Are you already running AB testing for your ad campaigns? If so, what trends or insights have you managed to gather? Let us know in the comments below!

Join the conversation


Submit a comment

Your email address will not be published. Required fields are marked *