Tracking Your Results: Types of Testing

By December 7, 2010 No Comments

Testing is at the center of improving conversion rates and should provide the foundation for what every marketing decision is based on.  The most common types of tests used to measure the effectiveness of your website are the A/B test and Multivariate test.

A/B Test

The simplest type of testing is the A/B test, which allows you to serve site visitors multiple versions of the same web page randomly and simultaneously to see which one performs the best.  The A/B test gives you the opportunity to measure the impact of design changes on key business metrics such as the conversion rates for a particular action.  Sometimes, versions A and B are competing designs served evenly between site visitors.  In other tests, version A could be your current web page and used as the control to compare to version B, which would be the experimental  edition and is served to a small percentage of users.  A/B testing is a great and relatively inexpensive way to measure the actual behavior of your customers under real-world conditions.  As long as the data you’ve collected is from a large enough sample size, you can conclude with confidence that if version A sells more than version B, then version A is the design you should roll with moving forward.

A/B tests are great for e-commerce and other sites that have one clear, ultimate goal, such as increased sales or leads.  Additionally, this goal must be measurable by counting simple user actions.  The major drawback to this method is the lack of behavioral insights you’ll glean.  While you will see which version gets better results, you won’t know why one version performed over the others without feedback from the actual users.

Multivariate Test

Another, more complex method of testing is the Multivariate Test, which allows you to analyze multiple variables of a web page at the same time.  This can help you save time since you won’t be limited to just testing one element at a time and you can still get accurate, meaningful results, without having to increase your total sample size.

Multivariate tests are often used to improve conversion as it relates to your call to action.  Any element a site visitor will have to interact with to complete the desired call to action you have for them once they arrive on your landing page is something you would want to run a multivariate test with.  When starting this test, you will be designing multiple combinations of the same elements surrounding a call to action.  These elements can include size, shape, color, and placement of elements.

This type of testing allows for almost unlimited variable combinations.  The only limitations are the amount of time it will take you to gain significant data to draw your conclusions.

David Lussky

David has been a Project Manager with Ninthlink since 2009. He enjoys the beach, electronic music, outdoor activities, traveling, and dominating in fantasy sports.

Leave a Reply