• February 4, 2014

The Basics of Display Advertising Split Testing Explained

Written by Ted Dhanik.

Testing is a term that gets thrown around in the world of display advertising, but the scientific aspect of marketing is very much a part of successful campaigns. Testing is a good method to confirm your suspicions and test the ideas you think you know. Through testing, you might find that you can double your conversion rate or reach new segments of your market. Without testing, your conversion rates are bound to stagnate. If you’re struggling with testing, or looking to implement new ideas, read on for some helpful tips to get started.

A/B Test

The basic A/B test pits two ads against one another in the hopes that one will convert better. The idea is that a marketer will take a page that is already working and look for methods to improve that page. Start with the easy stuff: your copy and your call to action, then move on from there. Begin with the copy from your successful ad and change one word. Use A/B testing to see whether your one word made a difference. Rinse and repeat, testing your theories one by one and keeping careful logs of your findings. Remember that an A/B test uses only ONE variable. More than one and your results get muddled.

Split Testing in Practice

The split test is made much more powerful with the proper tools. Google’s content experiments will do a lot of the heavy lifting for you once you have identified the URLs you want to use. The experiments send equal amounts of traffic to both pages, which removes unnecessary variables like time of day or traffic source from the equation. Once you have the experiment set up, let it run, even if you don’t like the initial results. The time frame will vary, especially if your new campaign is spending recklessly. You want to give the campaign enough traffic to have had a fair shot, but not too much.

Reviewing Analytics

Once the results are in, scrutinize every piece of data you have. Which pages are most popular? Which pages have the highest bounce rates? Which websites (placements) are sending the most traffic? Google Analytics is an excellent program for this, but there is much to be found in competitive research tools as well.

Implementation

Once you have some of these basic facts, you’re ready to approach your banner advertising with a bit of focus. For example, if you know that a high converting page also has a high bounce rate, you might want to improve the landing page by adding an exit pop to keep the user intrigued. Just remember to continue your testing, even if you find a verifiable winner. You can always improve!


Ted Dhanik is a marketer and business developer. As the president of engage:BDR, Ted Dhanik has specialized in lead generation for businesses in direct marketing. Find out more about direct marketing trends by visiting Ted Dhanik online.