PayU Reduces Checkout Page Drop-Off By Implementing Data-Driven A/B testing

PayU enables businesses across India to accept and manage payments online. As a payment facilitator, it is important that their checkout process be simple, intuitive, convenient, and not one that caused drop-offs for any reason.

Before getting onboard with VWO, Ravindra and Abhinav would conduct experiments based on their interactions with customers and merchants in order to continually improve their checkout experience. Based on the feedback of actual users, hypotheses would get finalized to improve the checkout experience. However, they faced two critical challenges:

• With the help of PayU’s dev team, Ravindra and Abhinav’s team used to implement A/B tests on PayU’s website at a server level. This cross-team dependency took a lot of their bandwidth, and didn’t provide them with an easy way to segment visitors based on their unique attributes such as location, platform, device type and more.
• There was no way for Ravindra and Abhinav to gauge how visitors were behaving on their checkout page, because all they had was hard numbers like time duration and bounce rate. PayU didn’t have a qualitative understanding of what exactly was happening on the checkout page.

The checkout page had been resulting in a number of drop-off cases, which was a cause for concern, as it represented lost sales and revenues.

What was required was a scientific way to test ideas and hypotheses related to the checkout page before deploying it.

VWO helped Ravindra and Abhinav in the following ways:

• As a first step, VWO provided them with the ability to easily segment the PayU website’s checkout page visitors based on a number of parameters such as location, device type and more, as they no longer had to implement the test at a server level.
• Second, VWO’s drag-and-drop A/B Testing Builder enabled them to drastically reduce dependence on its developers. By using VWO, the team could handle implementation of complex changes while increasing the speed of experimentation. Now, the PayU team could test more number of hypotheses per week compared to earlier.
• Third, VWO’s Visual Behavior Analytics tools such as Session Recordings and Heatmaps allowed the team to glean qualitative insights about their visitors’ behavior to be able to come up with strong hypotheses and better learnings from their test results.

The Test

PayU’s old checkout page required customers to enter their email and phone number before completing their purchase.

One of these two fields was essential, as local laws mandated that a final receipt had to be shared with the individual making the purchase.

Using VWO Form Analytics, Ravindra’s team discovered that entering both telephone and email address was becoming a source of friction, resulting in substantial drop-offs at this stage.

Based on this insight, they decided to A/B test if dropping the email address field would increase conversions. Here is how the two variations looked like:

Control

Variation

The results demonstrated that dropping the email field showed a statistically significant improvement of 5.8% compared to the control.

Based on the A/B Test results, the email field was dropped. This concluded that a number of customers did not have an email address, even though majority of them had a telephone number, this was further confirmed when PayU contacted a number of customers and they indicated the same.

VWO enabled improvement of the checkout process and delivered proven results through A/B testing. Since then, PayU’s goal has been to continually test and improve its checkout page and keep the user experience fluid while doing so.