5 Tips for Better A/B Tests

Every resource and asset you create - be it an ad, landing page or email, always consists of multiple elements. The design, copy, CTAs, headlines, ad placements… The list goes on. With so many factors influencing the performance of your creations, optimising them is definitely not an easy task. This is when A/B testing comes in useful - and here are your 5 tips for better A/B tests.

What are A/B tests?

A/B testing is a very common and effective way to optimise any digital asset you create. It revolves around two variants – A and B – which are both tested simultaneously to find out which is performing better.

How are the tests performed, you ask? Two variants of a landing page, email or ad, as an example, are shown to users at random - which means that 50% of your web visitors or email recipients will see the version A, whereas 50% of them will see the version B. Then, the A/B test will indicate which version proved more popular among your audience based on specific metrics of your choosing.

This might seem like a quick fix - but in reality, it’s not. It takes time and diligence to get proper results, and comparing the winning variant with a new variant on and on, to keep optimising your assets.

What are the benefits of A/B testing?

Essentially, A/B testing takes the guesswork out of your optimisation efforts. You can easily check the preferences of your users, and see how they react to your assets or any changes to them you might have in store.

What exactly can be achieved with A/B tests, then?

  • A/B testing helps you figure out what words, images, button placements, and other elements seem attractive for your target audience. This way, you can truly meet their expectations when creating any digital assets and resources - and even the simplest changes and minor adjustments can work wonders here
  • A/B testing helps you reduce bounce rates. In many cases, the assumptions website owners have about their target audience are completely wrong. As a result, web visitors and email recipients often “bounce” from the site or messages without spending any time checking them out. A/B testing can help you verify such assumptions and truly get through to your target audience. A/B testing helps you reduce bounce rates. In many cases, the assumptions website owners have about their target audience are completely wrong. As a result, web visitors and email recipients often “bounce” from the site or messages without spending any time checking them out. A/B testing can help you verify such assumptions and truly get through to your target audience.
  • A/B testing helps you increase conversion rates. This is basically what optimisation is all about. With A/B testing, you’re seeing for yourself what works and what doesn’t for your visitors and email recipients, which makes it easier for you to craft your layout, copy and visuals to achieve more conversions. A/B testing helps you increase conversion rates. This is basically what optimisation is all about. With A/B testing, you’re seeing for yourself what works and what doesn’t for your visitors and email recipients, which makes it easier for you to craft your layout, copy and visuals to achieve more conversions.

Once you’re convinced they work - here’s how to make the most of A/B tests.

Tips for better A/B tests

  1. Start A/B testing with no assumptions whatsoever.

    When you’re creating an asset and putting it to the test, you should have no assumptions as to how your target audience would react to certain versions of it. Let the users tell you which option works better for them - and don’t try to act against their opinion.

    The results of A/B tests might not be what you expect, and this happens more often than you might think. The key to making the most of them, though, is to draw data-driven conclusions and strive to optimise your asset accordingly.

  2. Test only one thing at a time.

    It’s true that A/B testing is all about comparing two versions of a web page, email, or any other digital asset. The thing to remember, however, is that these versions should have only one varying element. So, to give you an example - if you want to test a specific CTA, you should create two versions of the same page with only this specific CTA changed. Otherwise, it might be difficult to say what actually appeals to your users and what doesn’t, and attribute the results of an A/B test to particular elements.

    But, what if you want to test multiple elements of your asset (as you probably should do anyway)? Write down a list of such elements and prioritise them, it’s simple as that. Then, run your A/B tests - one at a time, to get a clear picture from the results.

  3. Put your trust in qualitative analysis.

    The best way to test anything is to ask your audience for detailed feedback or, even better, to see for yourself how they use your assets. Even if you consult the designs with your team or friends, you’re already gaining qualitative insights. Still, when you hear about a certain issue from your users first-hand or actually watch someone navigate your site - you get a strong proof of what might not be working as well as you expected.

    That’s precisely why user session recordings & replays work so well with A/B testing.Session replays are recorded visits of users actually browsing your site. Of course, since it’s basically an observation you’re making, it’s subject to your own biases. Nevertheless, recordings are a goldmine of information and a great tool for finding bottlenecks in your user experience.

  4. Use the right tools.

    If you want to do A/B testing the right way (especially if you want to leverage session replays), you’ll need a proper tool for this purpose. There are plenty of factors that affect such tests, which is why trying to conduct them all manually is really difficult (and frankly, often ineffective). Tools like LiveSession, Omniconvert, AB Tasty or Freshmail can easily help you create the right environment needed to run A/B tests. This, in turn, helps you get valid results that will make a real difference to your optimisation efforts.

  5. Define what metrics you’re after and document all your results.

    Before you run any test, decide how you’re going to measure its success. In fact, there should be one metric in particular that will help you determine the winning version of your asset. Such metrics can be different, depending on the actual result you’re trying to achieve. Pick something measurable which has an impact on your business. The conversion rate is a great example here. Other metrics that are worth tracking include bounce rates and exit rates.

    Once you have the metrics covered, don’t forget to document all your tests and their results. This will not only save you some effort on repeating it but also help you see the improvements over time. Not to mention that these results can reveal deep insights about your customers and their preferences, which is definitely something worth to keep in mind at all times.

Take your digital assets to the next level with A/B tests

There’s no single best practice when it comes to A/B tests - you can pretty much test whatever elements you want to. It makes most sense to test on pages that already have some traffic, as this is more likely to make a difference to your company’s bottom line. The most important thing, however, is to stay consistent and have patience.

Usually, a recommended time span for an A/B test is between one to four weeks. The recommended minimum is 7-14 days. This amount of time can be flexible, though, depending on the amount of traffic or interest your digital assets receives. The frequency of tests also varies - but keep in mind that the goal is to optimise your assets constantly.

For best results, combine quantitative and qualitative insights to come up with even better solutions for your digital assets. Once you’re at it - definitely give session replays a try. LiveSession is here to help you out!

Want more knowledge?

Get more tips and insights on UX, research and CRO. Zero spam. Straight to your inbox.

Share
profilePicture
Kalina Tyrkiel
Content Specialist @ LiveSession