Matter Solutions Logo (Simple Version)
April 24, 2017

The Importance of A/B Testing

Published: 24 April 2017 

A/B testing is a fantastic way to ensure you’re getting the most out of your promotional campaigns. After all, what is the point of running a campaign if you’re not achieving any results? With A/B testing, you’ll be able to ensure you achieve a higher ROI.

So what exactly is A/B testing, how do you create a test, and why is it so important?

What is A/B Testing?

Source: LinkedIn

A/B testing, which is also known as split testing, is the task of comparing two versions of a webpage, ad or app - whatever you want - to see which performs better. By changing elements, you are able to see what performs better and achieves more conversions. This takes away any guesswork, and you’ll be able to run a promotional campaign that you know will achieve the best results.

How exactly does it work? Once you have modified your ad, app screen or webpage, half of the audience is shown the original, while the other half is shown the modified version. With an equal amount of traffic being directed to each page, you can then measure the visitor's engagement levels. You can then see what impact the changed element had, and whether it is positive or negative.

What to test

There are many elements you can test in your A/B Test. It is important to note that you don’t have to test everything - or else you’ll never finishing testing! Instead, focus on what elements you think will make a difference.

Elements which you should consider testing include:

  • Headline - the words, size, and colour
  • Call to action - the text, the location, and colour of the button
  • Images - if using any graphics, consider altering to see any changes
  • Copy - change the description to see which one achieves better results

Creating an A/B Test

Here is a helpful framework which can help guide you through your next A/B test:

  • Know what you’re testing: are you testing ads, sales emails, web pages or apps?
  • Data: use your data to find out what isn’t performing. Do you have a webpage that has low conversions? You can then A/B test this page to see what elements cause the low conversions, and what alterations will make this rise.
  • Set goals: how much do you want your conversions to rise by? Setting goals will help you to measure whether or not the changes have contributed to it being a success.
  • List of variables: make a list of what variables you want to test. Make sure you don’t test too many things at once, or else you won’t be able to tell exactly which element is responsible for the rise or fall in conversions. Consider prioritising your list of variables, and how difficult it is to make a particular change.
  • Change variables: use your chosen A/B testing software to change your variables.
  • Undertake test: once you have the two versions, you’re ready to undertake the A/B test. Visitors will be split 50/50, with half seeing the original and the other half seeing the altered version.
  • Results: once the test has been completed, it’s time to check your results and analyse the data. You will be able to see which version performed better. If you haven’t made dramatic changes between the two versions, it should be easy to know which variable made the difference.      

Don’t panic if your new version didn’t achieve the results you wanted. Sometimes, you’ll end up with a negative result, or no result at all. Take away any lessons from this experiment, and try again with a new hypothesis.

Metrics to measure

Once you’ve run the A/B test, you’ll want to monitor your results. What should you be looking for in your data? Metrics to look at include:

  • Conversion rate: this is often the most important metric for businesses. You want to measure every completed conversion per unique visit.
  • Time spent on page: measuring how long people spent on your page will give you an insight into how interested your visitors are to your product or service.
  • Bounce rate: it doesn’t matter if you’re generating high traffic rates if everyone is leaving the website right away. If you have a high bounce rate, it’s time to investigate what the problem could be.
  • Unique visitors: this will show you how many new people are visiting your web page.   

SEO impact

What impact does A/B testing have on SEO? If you’re undertaking an A/B test incorrectly, whether intentionally or not, it can have a negative impact on your SEO. Google has released a statement regarding A/B testing which outlines ways to ensure your test has a minimal impact on your website. This includes:

  • No cloaking: you will be violating the Webmaster Guidelines if you show a different version of your website to Google’s crawlers. Manipulating the Webmaster Guidelines can negatively impact your SEO, and see your rankings drop, or your website will be completely removed from search engine results.
  • rel=“canonical”: to help Google know which page is the original, use the rel=”canonical” link on your alternative pages. This allows search engines to know the original page is the preferred version. This is a far better option than using a noindex meta tag.  
  • 302 redirects: if your A/B test involves redirecting users to the alternative URL, it’s best to use a temporary 302 redirect, rather than a permanent 301 redirect. This allows for the original URL to remain in an index.
  • Test time: nobody will have the same test time, and some tests take longer than others. If a test is being run for an unnecessarily long timeframe, Google may take action against the website.

It is clear that A/B testing is a crucial part of any promotional campaign. It will help ensure you run the best version of your ad, web page or app, and can achieve the highest results.

Ben Maden

Read more posts by Ben

Leave a Reply

Your email address will not be published. Required fields are marked *