A/B testing web design | Creative Freedom

Otherwise known as split testing, A/B testing is commonly used to compare and contrast different versions of a design side-by-side in order to work out the most effective strategies for branding, website optimisation, and most importantly, conversion.

Many A/B tests tend to focus on comparisons between single variables at a time in order to establish which is most effective. It can be used for social media, emails, mobile apps, landing pages, digital advertisements, and sites. So why use A/B testing?

The applications of A/B testing

A/B testing focuses on the actions of users in relation to a given design. In this way, it aims to demonstrate which designs and techniques resonate the most with users and potential buyers. A/B testing is most commonly used to test:

Page copy
Navigation elements
Visuals
Personalisation
Calls to action
Content
Sign-up forms
Pricing

This opens up a lot of possibilities for marketers and designers alike. It boosts engagement by demanding increased, transparent communication between users and developers. By testing across different groups of customers, developers can improve the efficacy of a campaign and personalise the right elements for the right users.

Furthermore, it provides you with a huge amount of data about user behaviour—the more testing that is run, the more intuitive the design process can become, ultimately reducing long-term costs and overheads, making your web-based campaigns more efficient:

“A/B testing, done consistently, can improve your bottom line substantially. If you know what works and what doesn’t, and have evidence to back it up, it’s easier to make decisions, and you can often craft more effective marketing materials from the outset. Just remember to keep testing regularly, since the effectiveness of anything can change over time.” – Cameron Chapman for Kissmetrics

When should I A/B test?

A/B testing is one of the most effective experimentation and assessment tools available—so when should you start A/B testing? This excellent infographic, produced by VWO.com, outlines a variety of scenarios in which A/B testing should be deployed. These demonstrate how A/B testing can act as a check on unsubstantiated assertions and proposals by challenging them directly—meaning that A/B testing not only allows you to improve your designs and campaigns, but also allows you to means-test decision making processes throughout your organisation. Some of these scenarios include:

When you are making decisions based on the Highest Paid Person’s Opinion (HiPPO)
When pro-innovation bias is being exerted
Instances of groupthink (‘bandwagon effect’)
When market research says something should be done in a certain way
When something works for your competitor
When you have little feedback, or only one suggestion, to act on

In our next post on A/B testing, we’ll be looking at how to start planning and developing an A/B testing procedure with a focus on drafting A/B hypotheses.