The What, Why and How of an A/B Testing Framework
20 / 08 / 2021

The what, why and how of an A/B testing framework

Understand how important an A/B testing framework is for generating more qualified leads and increasing user engagement across your website!

What is A/B testing?

A/B testing is a comparison of two variables at one time, also known as split testing. It allows you to see which variable performs best, for example, you can run an A/B test on your landing page with two different header images and identify which version generates the most conversions. 

 

Why should you conduct A/B testing?

By comparing two different marketing variables in an A/B test, you are able to better understand your audience, how to improve traffic to your website and generate more engaged leads.  

You’ll learn how certain elements of a users experience can impact their behaviour, and use A/B testing on an ongoing basis to continually improve this experience as well as increase a specified goal e.g conversion rate. By testing one element at a time, you are able to pinpoint the exact factor that demonstrates measurable improvement. 

 

6 Steps to Run an A/B Test

    1. Collect current data. Use analytics tools to gain insight into where you currently see a higher volume of traffic to your site, as well as pages that have low conversion rates or high bounce rates. 
    2. Identify key conversion goals. You need to know before you begin how you’ll determine whether the variable is more successful than the original. A conversion goal can be anything from a click-through rate to an email signup.
    3. Create a strategy. Once you know your goal, you can then assess what variants you want to A/B test and hypothesise why you think they’ll improve conversion rates. Variation A will be your ‘control’ (the original) and variation B will be your ‘treatment’.
    4. Use A/B testing software. This is the best and easiest way to monitor your test, as these tools usually feature a visual editor and can track results. For example, HubSpot allows you to duplicate an original landing page (the control) and then create an optimised version (the treatment) which you can easily modify.
    5. Allow your test to run. As visitors participate, you’ll be able to measure, record and compare how each variable performs. Ensure your visitors are equally and randomly split between the two variables to ensure fair results.
    6. Analyse your results. Once your test is complete, you need to review which variable performed the best and if there is a statistically significant difference. This will then help you decide if you should make the change.



5 Key Things to Remember Before Starting an A/B Test 

    1. Only conduct one test at a time. You’ll muddle your results if you carry out multiple tests across assets, offers and pages simultaneously. How will you know which test is resulting in conversions? 
    2. Same goes for variables - stick to one at a time. If you start to change more than one element on any given variable, you won’t be able to distinguish which change is driving conversions. For example, if you A/B test on your audience targeting and then also A/B test the colour of a CTA.
    3. Start big and then make smaller tweaks. A/B test your entire page with different variables and then start to hone in on the smaller details like CTA colour and image placement. Design two completely different pages with variable imagery, layout and style, then test them against one another. 
    4. Simple changes can have a significant impact. Once you have a clear winner for the overall page, then you can start to A/B test elements like colour, positioning and format. Even minor aesthetic changes and copy alterations can improve how your audience interacts with your page.
    5. Run your A/B tests simultaneously. Timing can have a huge impact on the performance of your test, so you need to run both variables at the same time to ensure your results are based on the variable element, not the time you’re running the test.

 

What to track when looking at your results:

  • Click-through rate 
  • Conversion rate
  • Bounce rate
  • Leads created
  • Demo requests
  • Sales

A/B testing will help you to better understand your audience and allow you to create a
set of optimised best practices that can then run across all of your marketing campaigns,
resulting in a better ROI. Not sure where to start? Let us know what you're struggling with and our experts will be happy to help.  

 

Header image source: Unsplash

20 / 08 / 2021 -

Prodo Insights

Need a new website - Download our Website Project Profile Template to get you started