Unibuddy’s chat platform redesign
Redesigning Unibuddy’s chat platform to increase sign up conversion by 23%
When
Mar 2024 - Jul 2024
Product
Unibuddy’s Chat Platform (web widget)
Team
Myself, PM, tech lead, squad of four devs
Janina
Start with impact, and before/after —> then go into how I did it.
INTRODUCTION
One of Unibuddy’s most powerful ways to show ROI (return on investment) to its customers, universities, is through the number of sign ups and interactions that they have on our platform. Indeed a few of our customers had churned due to low engagement. A large part of customers who struggled, either had low engagement on their websites, where Unibuddy gets embedded, and how the majority of traffic find Unibuddy widgets, or had not surfaced Unibuddy and its various traffic drivers in a way that drove a lot of traffic to thew page Unibuddy lived on. However, on the product side, we could see that of those that do land on a page with the Unibuddy widget active, about 55% drop-off without any interaction at all, and only about 6% of ended up signing up. Our aim was to improve these conversion metrics.
RESEARCH
Exploring the data
My first step was to explore the data. Understand where users were dropping off throughout the funnel. What hurdles existed throughout the journey, what interactions did users perform most regularly, and which of those interactions actually led to sign ups. The main things that I found were…
I looked at the data and it showed me these three things:
This
This
and this
I used these methods for research because, and with them these were my top takeaways
Student interviews
I also wanted to interview students to understand more about what their impressions were of the platform, as well as comparing ourselves directly to our competitors platform, to understand if they were doing anything that especially resonated with students that we were not. I performed both moderated and unmoderated interviews.
Findings - (“I don’t care about this picture”)
Tools used
Survey
Given that we had a sense from the interviews that one of the sticking points was the information on buddy cards, I sent out a survey to understand what the most important pieces of information were to students when making their decision of who would be most helpful to start a conversation with. The results showed, somewhat unsurprisingly, that there were differences between international and domestic students, and for students applying to postgrad and undergrad.
IDEATION
Workshop
Given all the information that I now had, I played it back to the team, and ran a workshop to start to ideate solutions. The structure of the workshop was… The results of the workshop were…
Experiments
To understand our users further, we ran a set of experiments.
Hypothesis: By decreasing the amount of choice and making the value more clear we expect more prospective students to take action.
At first all we want to drive is an increase in the numbers of initial actions taken. Our hypothesis is that by increasing the amount of initial actions taken it will increase the number of users who sign up.
Goal: Although we hope to eventually increase sign-ups, in this iteration we want to learn what percentage of users land on this experience with a different goal in mind. And does limiting the amount of choice available actually lead to sign ups.
The results of which told us…
Triage experiment
Next steps
After discussions with the team, including product leadership, we decided that our focus should be on the landing experience. We thought, if we could get as many users into the top of the funnel, we would aid the overall conversion the most. Given what we heard in the interviews, we decided our product could use a face-lift.
SOLUTIONS
Redesigning the ‘buddy card’
Building for the future, and the good of the system
inc. Before/after
Buddy card audit
