A/B Test

A/B Test

aka: multivariant, bucket or split testing is a series of scenario-based analytic methods that uses a baseline control sample that is compared to several single-variable test samples with the express purpose of improving and maintaining optimum conversion rates over a defined period of time.  These testing scenarios are most recently used to improve direct mail, and email marketing, as well as, the effectiveness of banner ads and landing page content. This data is most generally used for e-commerce sites that require a baseline for purchase statistics and even take into account drop-off rates, which in many cases can also represent unpredicted sales.

By implementing A/B testing on website elements, such as copy content, flow of layout, along with images and colors it is possible by reviewing a variety of test results to determine the improvements or variations that consistently provide the greatest benefit.

By using a control module, multiple samples of a test can identify and prove a desired outcome, as long as an audience of a sufficient size is used to detect significant and meaningful differences between the control and the various modules used. This method of control versus variables is sometimes referred to as a designed experiment and used widely among the largest and most successful Internet giants, which have proven to a great degree to maximize profitability over the relative long term. Some of the most notable are:

  • Amazon.com
  • BBC
  • eBay
  • Google
  • Microsoft
  • Netflix
  • Playdom
  • Zynga

On a small scale, most organizations can run simple marketing scenarios and compare results manually or without the aid of a robust computer program and scripting to evaluate their marketing strategies. However, for large corporations and e-commerce companies, the smartest and most efficient way to A/B test is to implement sophisticated tools, designed to aggregate and correlate results that are much more reliable in nature. The most popular (to date) A/B testing tools are: Google Website Optimizer, an open-source tool that can be used as a free version or as a licensed application, Omniture, which tests and targets, and Personvize- the latter of the two being commercially supported and offering a broad range of features that allow for almost unlimited customization.

Start your iContact journey today

Engage, WOW, and Grow! Your next big win starts here

you have been invited!

Go live with the
Customer Success Team!

In this 30-minute session, we will:

Choose a session