How to Do A/B Testing: A Guide to Improve Your Website Marketing and Conversion Rate


A/B testing is a powerful way to refine and optimise your website, attaining data-driven gains that deliver consistent ROI. But how do you run an A/B test? And what should you even be testing?

Here, we provide a comprehensive guide to A/B testing, taking you through what it is, how it works and how it can help your site perform. We also provide practical tips on areas of your site which you should A/B test, with suggestions on how to refine and improve on-page elements.

Quick Links

What is A/B Testing?

A/B testing is the process of comparing two versions of a web page to find out which performs the best. Two or more on-page variants are displayed to users at random, with data showing which is more likely to prompt users towards a conversion goal.

A/B testing is a fundamental part of optimising a conversion funnel. It allows for the complete fine-tuning of a website, ensuring that buttons, CTAs, forms, content and layout point users in the right direction.

Why You Should A/B Test

There are several key benefits to A/B testing and the process can have a major impact on how well a website converts. Below, we look at the reasons why A/B testing is always a good idea.

How Does A/B Testing work?

A/B testing works by creating a second version of a webpage on which to modify set variants, such as a title or button. Then, one half of site traffic is shown the original and the other half is shown the new variant.

From here, analytics tools capture data about how users are engaging and interacting with each version. This gives a clear view of whether the changes had a positive or negative result, or made no difference.

When the test is complete, you can choose to implement the change if the results were positive, make further changes based on data and feedback, or retain the original page.

Of course, there’s more to A/B testing than simply trying new variants at random. The process requires a considered approach, with plenty of forward planning to address specific goals and metrics, and accurate analysis of results.

Below, we take a look at the basic A/B testing process to show how it can be used to refine and improve experimentation.[

What Should You A/B Test?

Virtually every element on a webpage can influence visitor behaviour and how they interact and engage with your content. Even factors which you may consider minor, such as subheadings, can affect engagement and make it more or less likely that users will convert.

Below, we take a look at the on-page elements which should be A/B tested to ensure that they’re optimised for conversions. This isn’t an exhaustive list, but should provide a framework for experimentation and a guide on potential improvements you can make.

Design and Layout

One of the foremost areas for A/B testing is the design and layout of your website. It’s critical that the layout of your site, regardless of its overarching objective, leads users to your desired conversion goals, and provides the key information they’re looking for clearly and succinctly.

Taking an e-commerce website as an example, the design and layout should do the following:

  • Provide clear information – is key product and service information on the homepage and landing pages apparent enough? And is it put across in an engaging, uncluttered way? Here, a mix of media, including copy, imagery and videos, can help bring products to life without overwhelming the user with reams of information.
  • Be simple and fun to engage with – regardless of what a website is selling, the transaction process should be clean, simple and – if appropriate – fun and appealing to engage with. Creating detailed customer personas can help you refine the transaction process, ensuring its fit for your users and meets their requirements.
  • Highlight relevant reviews – customer reviews are powerful conversion drivers, so these should be prominent and accessible, giving visitors total peace of mind that others have endorsed your products and services.

From an A/B testing perspective, it’s possible to experiment with lots of different layout variations, placing things like customer reviews, testimonials, contact information, buying buttons and imagery on different areas of the page to encourage greater engagement and conversions.


Navigation is one of the most important areas of a website to get right and optimise for conversions. How users get around your site can only affect their experience and how likely they are to convert, but it can also impact how well the site performs in organic search.

In our guide to writing a website requirements document, we discussed the importance of having a clear site structure with a small number of pages leading from the main homepage, in a hierarchical structure. This type of structure is ideal for optimising your site for conversions, ensuring that pages connect fluidly and marry with people’s expectations and requirements.

For A/B testing, variants you may want to experiment with include the placement of the navigation bar (vertical or horizontal; bottom or top of the screen), the buttons on the navigation bar, and the type of pages which you group together. For a site to work best, we’d always recommend placing similar landing pages and product pages within their own subsets. For example, a site offering European river cruises would ideally group all European river itineraries under one button in the navigation bar; it makes the most contextual sense.

Copy, Content and Formatting

Another area of your site prime for A/B testing and refinement is copy. How you communicate with visitors and the messages you convey remains among the biggest drivers of conversions, engagement and brand advocacy, so optimising and refining it through A/B testing is a powerful way to improve your site’s performance and encourage interaction.

Below, we highlight the key areas you should consider when A/B testing your website’s written content:

  • Titles, headlines and subheadings – page headings draw the eye more than most other on-page elements, so the words you use are hugely important. Experimenting with different heading variations can garner positive results, engaging users differently to steer them towards your designed conversion goal.
  • Tone of voice – how you write can have a major impact on your site’s performance, so use a tone of voice which appeals to your target audience. A/B testing different writing styles can pay dividends if you want to really engage with and speak to your visitors.
  • Formatting – consider how the copy on your pages is laid out; ideally, sentences and paragraphs should be short (no more than 25 words per sentence). However, you should experiment with different formatting variations to find the style which resonates best with your audience.


Featuring relevant imagery on a webpage is a good way to draw the eye, improve context and make your pages more visually appealing. However, some images undoubtedly work better than others at engaging users and encouraging conversions, so you should experiment with different variations to find a theme and tone which resonates and creates an emotional response.


Many sites rely on forms to collect information about prospective customers and users. But where a form is placed and how it’s formatted can have a big impact on the likelihood of users filling it in.

Forms are among the most prevalent elements that businesses A/B test. Tweaking how they look, what fields to include and the messaging they contain can have a major impact on the total number of submissions. Experiment with form length, field type, messaging and placement to maximise the number of users who fill out a form and complete a conversion goal.

If you lack the resources to formulate, run and maintain regular A/B testing on your website, the Banc conversion optimisation team can help. Our optimisation strategists combine industry best-practice with advanced analytics processes to refine and test every element of your website, maximising ROI. Click here to learn more or call our team today on 0345 459 0558.

Avatar for Anita Pun
Anita Pun