Case Study: EtonX

EtonX is an online education service that offers students a virtual classroom experience. Its operating model (at the time) was B2C and B2B. It would sell individual purchases of online courses via its marketing site, and it would also sell block bookings to institutions around the world for groups of students to attend online. The courses focused on students achieving a more significant understanding of soft skills, such as public speaking, to further qualify students for pre-university application.

Online course purchase conversion

Objectives

• To increase online course purchases for B2C customers
• Reduce the time that customer service staff spent dealing with 'lost' users trying to purchase or register for courses

My first project was to look at the signup flow. The team had feedback from customers for various pain points they had. First, I mapped the entire series of flows end-to-end, then walk through with my stakeholders.

There were some opportunities from doing this exercise, points in which we collectively agreed that we could reduce some steps and make changes. These would reduce the decision-making and data entry points - barriers - to getting users through and paying for a product. The solution for this was pretty complex as the routes potential users arrive from are varied.

Data is also key to understand where customer pain points are, with drop offs, impression rates etc. In this instance the data was skewed (which is often unfortunately, all too common) as the marketing site and sign-lows were split across domains. The only way to measure customer problems was complaints channels. The customer service team spent 90% of their time dealing with issues from parents or students who couldn't purchase a course online or register for a pre-paid class.

Problem Statements

The following steps were to identify our problem statements, constructed to give you hypotheses in which you can test and further validate your assumptions, as well as help you prioritise what you want to go after first. As you can see here from this example, some of the biggest problems were that pre-issued course voucher codes were hard to enter throughout the sign up process, which is a hang-over from B2B selling routes. Much of the reason why the sign up and purchase journeys were problematic were because they were conflated.

Other things that we discovered throughout the process were that typically courses were being sold to parents and students in countries where English was often a second or even a third language. There were a plethora of other issues to consider also such as technical constraints with school networks, firewalls etc. This all came out through testing.

Proving our statements with testable solutions

I designed a series of new journeys to cover a new proposed solution which was to split clearly students and parents registering with prepaid courses and those who were prospective new students or parents.

I designed the new screens and flows in Figma, and then leaning on my rusty development skills I built the flows into a live working microsite using Node, MongoDB to capture (and delete) sign ups/fake purchases. The reason for doing this was so that we could move quickly, and that we can track efficiently the results of testing using things like HotJar.

I recommended that we did a split of qualitative and quantitive testing of the solution. So we devised a series of 1:1 sessions with parents and students with existing and prospective users. I used Typeform to build in a feedback loop into the prototype to capture feedback from users.

This lead to the refinement and release of a new marketing website for EtonX, it reduced the amount of time spent by internal staff, with a reduced amount of requests from staff and students. There was also an increase in course purchases from B2C customers by 14%. This work lead to working on other stages of the journeys throughout the product, such as the student dashboard, and the virtual classroom itself.

Virtual classroom

The following project was also to look at the actual Virtual Classroom.

Objectives

• Decrease the number of students failing to connect to classes through technological reasons
• Increase the usability of the virtual classroom to make it easier for students to use, therefore happier to use the tool
• Increase ease of use for teachers to administer a lesson, therefore leading to higher pedagogy.

Problem Statements

The problems initially discovered through testing and observations were that tutors found it challenging to use all the tools comfortably when teaching, getting the correct views and perspectives working right. There were also some problem spaces around the break out room functionality.

We also had an opportunity to improve the overall user interface. We built and designed the new version and went through a lengthy testing process.

Here you can see a waiting room that we built and remote testing this with a tutor who - had loads of problems using it on IE in one of our first passes. (That pesky browser).

The new VC, alongside the release of the new marketing website, was rolled out in the late summer of 2021. Signups have grown, more units (classes) have been sold as a result of this work.

Further concepts that I also worked on were relative to the problems encountered when teachers online were giving lessons to pupils situated in the same classrooms. Students would often be able to see each other and distract each other giving the teacher little or no control over the classroom.