Why And How To A/B Test Digital Banners

Detractors might say that banner ads (just like a lot of other legacy media) are dead. But if you have a website, you know how crucial they are. There are plenty of reasons why almost  $5.9 billion were spent on banner ads in 2020 alone. Promoting pre-orders and events, cost-effectiveness, and brand building are only some benefits of digital banner advertising.

Today, we are here to discuss one of the critical aspects of banner ads for marketers — testing. Because intuition and experience can only take you so far, and the audience always has the last laugh. With a focus on the preferred A/B testing method, we are about to explore the whys and hows of banner ad testing.

Why A/B Testing Won’t Be Passé Anytime Soon

A/B testing or split testing, as you would know, is a method of testing two different versions of the same ad with two separate audience groups to decide which one works best. So why is this age-old method still so relevant for the format of banner ads?

Here are a few reasons:

• Understanding of Consumer Preferences: Despite focus tests, local surveys, and SEO research, no method is as insightful as evaluating consumer reaction to the actual communication. A/B can help you make the most minor decisions that matter, such as which font your banner should have?
• Understanding of Consumer Behavior: Say you have two variants of a product, namely chocolate and berries, of which you have been promoting chocolate more on the shelves. However, A/B testing has fetched better results for the berries variant. You might want to conduct product research and find out if consumers are actually demanding berries more.
• Elimination of Ad Fatigue: Consumers stop noticing an ad when they come across it repeatedly. A/B testing can help you pick their brains over what they would like to see and how often they would like to see it.
• Better Improvement Curve: When you make A/B tests a practice, you will gather patterns. These will assist you in making essential design and copy decisions for the long run. For example, you might notice that 2 out of 10 of your banners with pastel backgrounds have worked well. You could then shift to pastels altogether.
  
  
  

What Should You Be Testing?

A/B testing is no mean business. You ought to test one element at a time, and there are quite a bunch of factors. If you test multiple elements together, you wouldn’t be sure about which one worked.

Here are the key elements that you should be testing:

• Headline: Microsoft Bing, in 2017, was able to grow annual revenue by an incredible 12% by simply changing its display ad headline. Headlines are the first things that viewers notice. To test these paramount elements, vary the placement of headlines, play with fonts, sizes, colors, and power words. Remember that the point of a headline is mostly evoking curiosity.
  
• Copy: Who would have thought that changing a simple “Get $10 off the first purchase. Book online now!” to a “Get an additional $10 off. Book online now” would double CTRs? That’s precisely what happened with Straight North, a marketing agency. Copy can make or break an advertisement, even though it is not the first thing a viewer notices. So, let every word be put to the test and fight for its place.
  
• CTA: CTAs are the POS (Point Of Sale) of banner ads. The presence or absence of a CTA button, its color, copy, font, placement, etc., must be tested to ensure maximum conversions. Changing a simple ‘Buy Now’ to ‘Try Now’ can provide additional clicks on a new brand’s banner, for example.
  
• Visuals: Visuals are critical psychological catalysts in a banner ad. They are also one of the most challenging aspects for marketers. They can lead you to crucial decisions, such as conducting a photoshoot for your original products because stock images no longer work with your audience.
  
  Everything from backgrounds, fonts, and colors to the placements of these elements should be put to the test. Don’t hesitate to play around and experiment. At the same time, don’t forget to adhere to brand colors and style guides. A/B tested visuals can lead you to that perfect, tried, and tested banner ad.

• Timing: This is important from an economic point of view. Suppose you find out that your ads perform better on a particular day of the week and at a specific hour of the day or night. It will do you a lot of good to concentrate your budgets in that time frame. This timing would depend upon your industry, category, brand, and consumer profiles. You must also look at the time when searches for your brand/product/category are at a high and test your ads in those time frames.
  
• Landing Page Forms: The purpose of a landing page form is to provide value for the information you seek from consumers. To check if you’re providing enough value, you need to test the length, the position, the color, the design, and the content of your landing page form.

A/B Banner Test Examples

Here are a few examples of A/B tested banners found across the internet. 

LinkedIn: Here LinkedIn has tried two different banner sizes and formats for the same design, on the same page. Which one gets a better CTR will help them decide the size and positioning that works for this particular offering. 

Grammarly: In this classic example, Grammarly, the Grammar Nazi this world swears by, has experimented with static and dynamic banners, apart from altering the size as well. Which one do you think would fetch better results? The static, sleek one on top of a page, or the GIF in the middle of a page?

Dell: Dell has placed the same banner ad at two places on the same page, one on the left and the other in the center. Both are about one scroll away from each other. The only two differences are the size and the copy beneath the second banner. They have smartly tried to test if some extra copy outside the banner helps or not, apart from checking where a viewer’s attention is drawn first on the page. 

A/B Testing Banner Ads: Things to Remember

Here are a few make or break things that you must take care of while A/B testing your banner ads:

1. Ensure that the entire creative team (mainly copy and design) is on board with the tests. You don’t want one of those classic creative-strategy-analytics conflicts, do you?
2. Establish which KPI you are measuring. Don’t go shooting in the dark. 
3. Test only one or two aspects at a time. This needs reiteration as it is the very essence of A/B testing.
4. Decide the time for which you want to run the test. You cannot wait for too long or too short a time to know what works. This is especially relevant for time-specific or seasonal ads. It might so happen that you lose valuable inflows simply because you tested too much and didn’t run the right ad for a long time.

Takeaway

A/B testing can be a transforming exercise for your brand’s ad campaigns. However, you also need to realize that A/B testing is not quantum physics as extensive as it is. It requires, as any other research practice, a lot of trial and error. So our parting shot would be: be patient, be resilient, and happy testing!