Docstoc

How to correctly A/B Test with your online ads

Courtesy of Daniel Waisberg

Successful advertising campaigns are not random flukes. Instead, the most successful campaigns are the ones that are based on real research and honed in a process called A/B testing.

While the term A/B testing isn’t as popular as SEO or web analytics it is a process that was around long before the Internet.

What is A/B Testing?

In marketing A/B Testing or split testing is a method for comparing how individual elements in campaigns factor into how successful or well received it is by the consumer. This testing gives insight to consumer behaviors and allows an advertiser to adjust campaigns to return the highest possible conversion rate.

First used as a scientific tool for testing elements against a controlled sample, this method of testing was quickly adapted by direct mail campaigners to determine which marketing piece returned the most responses.

It has now been adapted to determine the success of online marketing. A/B testing is being used by savvy marketers to test landing pages, emails, and banner advertisements in ways that take most of the guessing out of the decision process and making the ROI or return on investment a much more profitable venture.

What should an A/B Test for?

What an A/B test is used for is determined by what the goal of the testing is. If testing to determine which banner gets the most customers to visit a landing page then that is the element in the data that is focused on. If it is to find out which landing page design gets the most users to subscribe to the email list then that is the element tracked.

There are a variety of ways to use an A/B test in marketing and each of these tests can give real insight into what customers look for. Even though each test is focused on a different element there are some common things that A/B testing are used to determine.

For example:

  • The wording, style, and placement of any call to action
  • Website layouts
  • Product pricing, promotions, and descriptions
  • Text quantity
  • Effective element of form’s
  • Images

Who uses A/B Testing?

The process of refining marketing campaigns through A/B testing is one used by some of the industries biggest companies. (Click on the name to read more about their market testing.)

What kind of tools are available for A/B Testing?

For the beginner there are several tools available to set up A/B testing. All of these tools are highly recommended and trusted. Just remember this is just a sample of what is available so researching is encouraged.

  • Google Website Optimizer offers their free A/B testing tool as a great option for anyone looking to get started right away.
  • Visual Website Optimizer is the self-proclaimed “World’s easiest testing tool” and they feature such a WYSIWYG editor that is point-and-click by design and does a lot of the work for you.
  • Adobes Test&Target offers the ability to visually create A/B testing while also allowing segmentation for better targeted tests.

Sample A/B Testing for a Banner Ad

To create a simple A/B test for a banner ad create two banners that are similar in all aspects except for one. The difference can be the call to action text, graphics, or layout. Now, place those ads like you normally would on Google, Facebook, or wherever you exchange them.

After the campaigns have had time to gather enough data compare to see which one achieved its goal better.

Sample A/B Testing for an Email Campaign

Create two identical emails except change one element. For this example change the time limit on the call to action from “limited time offer” to “offer expires on” a specific date a few days from now. Now, set up separate coupon codes for each version and send both off of at the same time, each to half the subscribers on the company’s email list.

Give the campaign enough time to saturate and depending on the goal of the email check the results. If one coupon code out performed the other that suggests a more successful campaign.

Be sure to do multiple tests before drawing any conclusions as timing can effect results greatly.

Sample A/B Testing for a Website

When wanting to test a website it is suggested that the best way to do this is by choosing one of the tools suggested above. These tools are designed to allow a website to test a single element or an entire redesign. They also provide a lot of valuable data collection.

Basically,testing two different designs of a website allows you to divide traffic between these two separate versions of a landing page and then measure how each one performs. Depending on what your testing for you can use the collected metrics to select the version that performed better.

 
;