How A/B testing can improve conversion rates

Today we're going to introduce you to A/B testing in PlusZero.

A/B testing is a way to randomly divide your website visitors into two groups, exposing one group to an existing webpage and the other to a redesigned webpage, and then quantitatively measure which of the two groups performs better to accurately assess how effective the new design is compared to the old. The success of the test is often judged by key business metrics such as signup rates, return visits, and conversion rates.

At Pluszero, we work with many different companies to optimize their business through A/B testing. We hope this article will help you find the right way to optimize your business by using data effectively.

Today, we'd like to introduce a case study on how PlusZero optimized its A/B test conversion rate.

How we improved the events page

The main challenge for this domain was to increase users' content consumption and conversion rates while decreasing churn. In this case, we mainly analyze what content users consume the most and how their key KPIs perform when they consume it versus when they do not.

When we segmented the data, we found that sessions that viewed the event page had an overwhelmingly higher conversion rate than sessions that didn't, making the event page one of the most important pieces of content to positively influence users to convert.

The events page is performing well, so the domain is sending most of its ad landings to the events page. However, due to the high percentage of sessions coming from ads, the exit rate on this page is overwhelmingly higher than other pages.

In this case, it would be desirable to have a window for users who came from ads to consume other content on the site, but there was no such window before. We hypothesized that inserting a GNB button at the top of the event page to consume other content on the site would increase conversion and content consumption rates.

Furthermore, when we analyzed the behavior of our converted customers, we found that they consumed more content and converted through the GNB button at the top of the page.

Therefore, we conducted an experiment with the existing event page as the original and GNB inserted at the top of the event page as the variation. As a result, the exit rate of users decreased and the conversion rate and content consumption rate increased significantly.

This experiment shows that an effective page plus a great customer experience creates an even greater synergy. For most websites, events and promotions are very effective. Using data to identify weaknesses and improvements within those pages will go a long way toward optimizing our business.

Example of a check out page improvement

On a company's website, users apply for broadcast internet service through a subscription confirmation page and an application form page.

However, 72% of users bounce from the subscription confirmation page to the next application page, and 88% of users bounce from the application completion page to the final completion page.

Having such a high percentage of customers bounce during the conversion stage is one of the biggest hindrances to a successful business. After analyzing the data, we found that the average dwell time and bounce rate on that page is high compared to other pages. In conclusion, we hypothesized that in order to increase the conversion rate of the website, the UI / UX of the Checkout 2 page should be improved.

Our first priority was that the Checkout 2 page had too much required information for users to enter and the page was too long, leaving a lot of room for users to bounce.

So, we broke the existing page into multiple pages and steps, and gave users a new page transition as they entered the required information.

This lowers the barrier to entry for users to enter their information and allows them to see their progress at a glance with a progress bar.

When we set the traditional scroll down to Original and the new page transition to Variation, we saw a decrease in user bounce rate and an increase in conversion rate.

If your website's checkout process is complicated and you're losing a lot of customers on that page, as in this case, simplifying the page structure will go a long way toward optimizing your website.

When making improvements to your website, formulate experimental hypotheses based on your data, and then A/B test them to see how they perform.