What is A/B Testing and How You Can I Use It?

What is A/B Testing and How Can I Use It?

A/B testing can help maximise the effectiveness of your marketing efforts by providing insights into a single change you make to a piece of content. As a result, it can be the difference between meeting a key performance indicator (KPI) or surpassing it with flying colours.

This brief guide will help you understand the basics and provide you with some tips on running some A/B testing of your own.

What Is A/B Testing?

A/B testing, also known as split testing, is a way of optimising content, whether it be content on a webpage, a marketing email or advertisement, and change a single element of that content.

It’s like a science experiment that you may have done at school. You make a small change and then collect data to determine if that change has been successful.

Black & Decker “Buy Now” vs Shop Now Test

Black & Decker has a well-known A/B test on its call-to-action (CTA) used on the product pages of the DeWalt website (owned by B&D).

They experimented with “Shop Now” against their original CTA of “Buy Now”. They discovered the original “Buy Now” label drove conversions to increase by 17%. Over a year, the projection is a six-figure difference in terms of the impact this seemingly insignificant change can have.

Black & Decker increase conversions by 17% WITH A/B Testing

However, this doesn’t mean that all businesses should use “Buy Now” as their call-to-action label on product pages. You should decipher what works best for you and your customer base since no two businesses are alike.

Types of A/B Tests

There’s not really types of A/B tests. As is usually the case with marketing, the only limitation is your ambition and creativity. If you’re not sure where to start, think of a small change, any change that you feel could deliver a result, whether positively or negatively.

It doesn’t have to be websites. It can be subject lines on emails or the call-to-action text on pay-per-click ads on Google. The advantage here is that the sample size is likely to be much larger, so you’ll be able to get a quick turnaround for meaningful conclusions. In turn, this will help to deliver some “quick wins” for your marketing and sales teams.

How Do I Run an A/B Test on My Website?

Step One: Devise a Hypothesis

When looking to carry out A/B testing, you must have sound reasoning as to why you’re doing it. In the Black and Decker example, the company believed its customers may have wanted a more definitive action to describe what they’re trying to do. “Buy Now” is the original text, and there’s no mistaking what it’s for, whereas “Shop Now” covers several actions.

 Another critical point is that it doesn’t just have to be minor changes in a webpage. You could add features like personalisation in a marketing email, the email’s subject, or the order of items.

However, the changes must be small. If the changes are too profound, it can be difficult to pinpoint what led to the improved or worsened statistics.

Get creative with others if you can. Take a bit of time to take inspiration from other websites, adverts or emails.

Step Two: Have Two Versions of the Same Page with One Small Difference

So, you’ve now got something to test. You’ll need to duplicate the page and make the small change you have a hypothesis about in Step One.

Your original will be your control version, which will need to remain unchanged, and the other will be your test version. It’s essential they are the same bar as one change; otherwise, you won’t know if that small change is responsible or if something else is instead.

Step Three: Choose a Timeframe for You’re A/B Test

Deciding on a time frame is crucial, so you have a good sample of data to ensure the accuracy of your findings. Otherwise, you may find the data is terrible, and the experiment can’t reproduce the same result.

How long of a timeframe will depend on the amount of traffic your website receives. Having thousands of visits is probably a good enough sample. But suppose your site’s traffic is only in the hundreds. In that case, you may need to run your experiment with two different landing pages over several months to get enough visits for a more representative data sample.

The bottom line is that you’ll only be able to derive accurate and meaningful conclusions from helping you achieve or surpass your KPIs when you have enough data to support it.

Looking for Help with A/B Testing?

XHost has a wealth of experience with A/B testing across search engine optimisation (SEO), pay-per-click (PPC), email marketing, and e-commerce. To find out more, contact us today.