A/B testing

Optimise Your Website or App Through A/B Testing:

1. What is A/B testing?

A/B testing involves the use of a controlled experiment where visitors are randomly allocated to different experiences to identify which, if any, design is statistically better at getting users to convert. Because visitors are allocated on a random basis to each variant, we can be confident that any statistically significant difference between the conversion rate of each experience is due to the design, and not other external factors (e.g. source of traffic). 

2. Why should you use A/B tests?

A/B testing is a scientific method for optimising digital experiences and improving customer outcomes.

The Benefits of A/B Testing:

1. Improved content as you can test to see what type of content is most likely to make visitors stay on your site or app for longer.
2. Enhance engagement by testing to see which elements users most interact with and so this will help reduce your bounce rate.
3. Minimise cart abandonment by testing your checkout journey.
4. Optimise landing pages by conducting redirect tests to identify which page is most effective at converting visitors.
5. Increase your conversion rate by systematically A/B testing key pages on your site or app.
6. Achieve a high conversion rate by applying learnings from A/B tests across your site or app.
7. Resolve disagreements about website or app designs by letting users decide through online experiments.
8. Develop a culture of experimentation to improve innovation and enhance the user experience.
9. Quickly obtain results by testing on pages with the highest amount of traffic.
10. Minimise risk to revenues by testing major changes before they are implemented in your live environment.
11. Increase sales and revenues by building a program of A/B testing.

3. Should you use A/B tests?

A/B testing can allow you to deliver significant uplifts in sales and revenues for your digital experience. Ideally, a website should have a high number of conversions a month (e.g. over 1,000) to ensure experiments have a good probability to deliver a conclusive result within a 4 week period. However, in the real world, it is better to run an A/B test to gather data to inform your decision-making than to make changes without any supporting evidence. 

In addition, you may have a high number of micro-conversions to begin the process of optimising your site. For example, if you have 1,000 add to basket events or registration completes a month, you can set this as your primary objective for an experiment. It’s not ideal, but it allows you to at least get more users further down your conversion funnel.

4. How can we help your business?

We manage and build A/B tests for clients to optimise their digital experiences using online controlled experiments.

Management, build and analysis of A/B tests. Plus, web analytics, visual analytics, competitor research and usability testing, to identify and develop hypotheses for A/B tests. 

A/B testing tools, such as Webtrends Optimise, randomly select users for different experiences and analyse their behaviour to scientifically identify the optimal design. 

With the launch of Google’s free solution, Google Optimize, A/B testing is now accessible for even relatively small businesses. Optimize allows you to run redirect tests to compare the performance of different landing pages and make changes to existing pages for iterative testing.

5. Why do we code your experiments?

It is better to manually code experiments than rely on the visual editors offered by many tools. Manual coding minimises the amount of script on your site and is less likely to have an impact on load speed. Visual editors often duplicate code as changes are made and are not recommended for best practice optimisation. 

Conversion Uplift follows industry best practices for experimentation, and employees maintain their skills and knowledge through the CXL Institute.