Create and test

Written in collaboration with Marc Miller

If you’ve ever visited a webpage on the internet, chances are you’ve been a part of A/B testing. At its basic level, an A/B test is a controlled experiment that tests a user’s interaction with a webpage by serving them multiple experiences of the same page to see which page resonates with them the most. Tests can be performed on any part of a website and can be as simple as changing a single button or headline of a webpage, or as complex as a complete redesign of a page.

A/B testing is nothing new

We use A/B testing every day as a way of determining what is the most impactful, meaningful and profitable way of doing business with our guests. For example, during a seat sale on, our digital team may decide to change the wording of the homepage or alter the look of a button from time to time. These are different experiences served up through A/B testing and in most cases, they lead to valuable insights about what our guests want and how we can give it to them.

In testing, data wins every time

We all know that words matter, and when it comes to attracting audiences and increasing sales, choosing the right words can matter a lot. If you’re debating between two lines of copy or the placement of a button, let the data determine what performs the best. The beauty of A/B testing is that it removes the guesswork, bias, opinion and hierarchy from decision making and provides data-driven, impossible-to-argue-with solutions for various problems. It can even go as deep as showing our guests entirely new experiences on certain pages or making incredibly subtle variations that lead to massive improvements.

The benefits of testing

A/B testing can be used to make money, save money, save time, improve the guest experience or even lead a guest in the direction we want them to go. Increasing revenue is perhaps the sexiest part of testing and making a seemingly small change to the copy on a page, or changing the colour of a button, can have a massive impact on guest behaviour and their willingness to make a purchase. It also allows our team to test an idea and prove it before we invest in it.

Of course, there are certain risks that come with testing. What if the new experience breaks? What if a guest dislikes it and doesn’t act how we thought they might? Creating and running fail-safe tests is part of A/B testing and we cannot be afraid to fail. Sometimes, there are more insights from failures than from successes.

Creating and running A/B tests follows a set of rules, just like any process. Ideas can and should come from anyone in the company; generating exciting, new and different ideas is what fuels testing. From there, the process becomes more simple where we create and run the test and, of course, analyze and report after it’s finished.

Not all tests lead to conclusive winners though, and some tests need to be run multiple times depending on a number of factors. This mentality means that A/B testing and optimization never really ends.

Nothing is ever finished, as it can always be optimized

The testing mentality is not only limited to websites and can spill over into all areas of a business. If a member of the recruiting team wanted to test a job posting, their objective could be to increase qualified applicants by writing two different job postings and determining which led to better candidates. The decision-making is then left up to the user and in turn, we cater to them in a way that we never have before, which leads towards a more data-driven, guest-focused experience for all.

This is the way that the most successful companies in the world are going, and it’s not too late for us to join them.