A/B Testing Explained

When you’re working on your marketing campaigns, it can be difficult to know in advance what’s going to bring you the most success. In a lot of cases, it might be trial and error, at least initially. A/B testing simplifies this task and defines what will perform best for your business.

Join us as we take a look at A/B testing.

What is A/B Testing?

A/B testing is the process of running two versions of the same element simultaneously, with just one variation. You can test elements on your website, emails and your ads, and the more testing you do, the more success your business will enjoy.

According to Invesp, 71% of businesses conduct A/B testing twice a month and 60% of companies consider it to be “highly valuable” for optimising conversion rates.

The purpose of A/B testing is to see which variant performs better, that is, gets you more clicks, sales, sign-ups etc. The possibilities are far-reaching and the end result will give you a clear idea of what drives your audience towards your intended goals.

To do it correctly, it’s vital that all elements remain the same, except for that one single variable. This means the same audience demographic and size, as well as the same time you’re conducting your test. Failing to do this will leave you unable to pinpoint the reason behind one outperforming the other.

Types of A/B Testing

You can A/B test a multitude of areas on your website, ads, and emails. Take a look at our list of things you can A/B test:

Website Colour SchemeColours impact how people feel, and by extension, how they feel about your brand. Therefore, it stands to reason that your choice of colour scheme for your website is a vitally important decision.

Call-To-Actions – When you’re encouraging your audience to click through to your blog, will your CTA button say “Read More”, “Learn More”, “Take a Peek” or something completely different? This may seem like a trivial aspect but how you word your CTAs will impact how many clicks you get.

Your Contact Forms – Short vs long, pop-up vs static, to the right of your page vs the left. There’s lots of things to A/B test about a contact form. Just remember to keep it above the page fold on your site!

Images – Some images will be more appealing and appropriate than others. Your audience demographic may be more inclined to like certain images, even if they don’t consciously realise it. Using different visuals on your website, emails or ads can alter people’s perception of your brand and therefore influence your conversions.

Ad or Email Text – Your ad and email text should be enticing, persuasive and accurate. It should let your audience know what to expect when they click through to your website. If you’re spending money on your ads and email campaigns, you want them to perform as well as possible to make sure you’re money is well-spent. A/B test different texts to see what wording and/or writing style gets you a higher CTR (Click Through Rate).

Email Subject Line – The subject line in an email is normally the deciding factor between opening, ignoring or deleting. Spend time A/B testing your subject lines in your email campaigns to learn the type of wording your audience prefers.

Before You Begin Testing

Before you begin anything, you need to determine what variables you are experimenting with and what your goals are for the test. Your goals could be to lower your bounce rate or to increase your email sign-ups. It’s completely up to you.

After this first stage, compile a list of A/B testing ideas and a hypothesis, detailing your expectations. Once the testing is complete you can see how close to the mark you were.

Consider how long you will run your test for, as well as the audience you’re targeting.

Let’s A/B Test

A/B testing is actually quite simple, with 63% of companies believing it’s not difficult to do, and only 7% believing it’s “very difficult”, according to Invesp.

To A/B test your website pages you can use a software such as Optimizely that allows you to create a new version of your webpage and run it alongside your existing one. The software will run your website and present different versions of it to different people.

While doing this, the software will gather analytics to identify which version performs better.

It’s worth noting here that A/B testing can increase page-loading time. Those hosting their own website on Windows WAMP for example, could run into speed problems. Ensure you have a good hosting plan in place before you begin. Slow-loading pages can impact the results of your tests, as well as your SEO.

To A/B test your PPC search engine or social media ads, you simply create your ads and run them at the same time.

To A/B test your emails, just send different versions of an email to different contacts.

When You’ve Finished Testing

Don’t be tempted to end the test before it’s pre-determined run-time, even if it seems like there’s a clear winner. You need to leave it alone and allow it to happen. Otherwise, you could end up with inaccurate results.

After your A/B testing, analyse your results carefully. You can then go on to A/B testing the winner against another variant until you’ve discovered the perfect combination of website, email and ad elements.

Final Words

A/B testing is a powerful method designed to increase the performance of your marketing and website materials. In a nutshell, it’s all about figuring out what pushes your audience to take a specific action. This could be a tweak in your colour-scheme or a change of wording to your call-to-actions. When conducted carefully and accurately, A/B testing can enhance your conversions,