Make AB Split Testing Decisions Based On Complete Data

Make A/B split testing decisions based on complete data

The process of refinement towards a perfectly optimised website is continual voyage of discovery for many user experience teams. As more effective methods that enhance the navigation of a website driving the lead to enquire are found – then the process continues. But as we’re all aware, what’s successful for one business is not necessarily successful for another. With different demographics, products and services there really are very few generalisations as to how to turn a website into a persuasive sales and marketing machine.

In recent years, the need for A/B testing has become a popular method to discover how your leads and customers respond to minor changes on a single page. By creating a variation against an original web page, teams can determine whether it produces a positive result or not. It’s commonly used as a way to validate for example a new design element before the change is rolled out across the website.

To make this testing process easier, two former Google employees developed software that would, “provide a platform for businesses to be able to conceive and run experiments that helped them make better data-driven decisions.” The result is Optimizely, an easy-to-use interface that lets users create a variation and monitor the results.

If you’re testing a web page that prompts a lead to contact your business by both a web submission form and by advertising a phone number, then it’s essential that you not only track online enquiries but also telephone enquiries. This is where Delacon’s call tracking solution provides that all important software piece that will enable you to analyse all data. If you only track online contact forms then you risk making decisions without complete information.

As A/B split testing is essentially a data driven exercise, it is imperative that you have all the appropriate data available.

This is all the more important because phone leads generally attract higher revenue than online web form enquiries. Therefore in a fictional situation, the original version of a web page may have more web leads and fewer calls – but the variation attracts fewer web leads and more calls. Under this scenario, the variation could be considered more successful because the overall revenue generated would be higher. Without the inclusion of call tracking, it would be extremely difficult to accurately determine the results.

With Delacon’s call tracking solution, you will be able to analyse the results of an Optimizely A/B test in Delacon’s reporting tool, Optimizely and in Google Analytics. In this way, you can also take into consideration secondary and tertiary data that may also play a part into determining which variation is the most effective.

Lyndon Barnett

Lyndon has always been fascinated by technology.

Leave a Reply

Your email address will not be published. Required fields are marked *