Unibuddy Community

How “Student Matching” increased conversion by +20% for universities

Overview

Product

Unibuddy Community is a group messaging platform that enables Universities to connect with offer holder students, to provide support before enrolling and help students foster a sense of belonging to the university whilst also increasing yield (yield = students who convert from an offer holder into an enrolled student) for universities.

This case study will walkthrough how universities have achieved an average of +20% increase in conversion through Unibuddy Community’s “Student Matching” feature.

Role

Lead Product Designer

Outcomes

+22% increase in undergraduate conversions

+33% increase in all messages sent in Community were between matched students

50% of all direct messages were sent between matched students

Problem space

The following problems have been identified through customer and market research.

💰 Business (Unibuddy) problem

Since Community’s MVP release in September 2021, the business is observing low MAU and engagement levels across the platform. The Business would like to identify a USP which will differentiate them from competitors to increase ARR from attracting new business, increase retention with existing customers and boost user engagement.

🏫 Customer (university) problem

Universities are facing challenges of sustaining engagement with prospective students before they enrol - also known as “Summer melt” - which is negatively impacting conversion.

👩‍🎓 User (student) problem

Students main worries are whether they will be able to foster a sense of belonging at the university and fear they will not be able to make friends.

Discovery

The Product Manager and I conducted semi-conducted interviews with 12 offer holder students and 12 university staff members to dig deeper into the problem space. In collaboration with the marketing team, we also sent out a mass survey to further validate the problems we had identified through qualitative research.

Findings

92%

of offer holders say connecting with incoming students is critical to their journey of enrolling. Their number one worry is whether they will make friends at university and foster a sense of belonging.

95%

of university administrators say peer-to-peer connections are critical to students deciding to enrol at University but they are unable to facilitate engagement as they are time poor and low in resource.

Additional insights

Per persona

Offer holder students

Offer holder students are adults who are waiting to enrol at university upon accepting an offer to study.

💡 It’s difficult to sift through the online noise and find authentic information about the university. Students need authentic insights to help them envision what life will be like at university, and to help instil confidence in their decision

💡 Offer holders often feel nervous and apprehensive to chat with university students in Facebook and Whatsapp group chats. But being part of a chat helps students to feel like they are supported prior to enrolling

💡 Students need answers to their concerns/queries before arriving at university to help reduce worries and anxiety. Those in the US typically end up withdrawing their offers as they cannot figure out how to afford university fees

Enrolment staff

Enrolment staff are often university recruitment members or admins who use unibuddy to conduct recruitment activities.

💡 Admin teams are low in resource and time poor, particularly after the pandemic

💡 Universities need more data about the offer holder students to help inform recruitment activities

💡 Universities are struggling to engage students which negatively impacts student yield and conversion

Opportunities

As a Product trio (PM, EM, PD) and in collaboration with the VP of Product and CEO, we analysed the discovery insights and identified key opportunities for Unibuddy Community in the form of JTBD.

👩‍🎓 Facilitate connections

Help students to find other like-minded individuals looking to enrol at the same university, to help form a sense of belonging and reduce anxiety over forming friendships

👩‍🎓 Encourage curiosity

Enable students to ask questions which can be answered quicker than an email, and with authentic insights

👩‍🎓 Inform & Prepare

Inform students of application and registration deadlines to ensure students feel confident they have everything in order before enrolling

🏫 Enable action

Provide universities with data insights from Community to inform them of students who are not engaging and pose a potential risk to not converting into an enrolled student

Cross-functional collaboration

The opportunities we identified from our discovery were translated into HMW statements and taken into cross-functional ideation sessions. I facilitated workshops with various members across the business - Product Marketing, Developers, SME’s, VP of Product, CEO - to encourage diverse thinking and generate hypothesis to test, to help solve our key problems in Unibuddy Community.

Ideation outcomes

The following ideas were considered the “top voted” by a collaborative team, to help students foster a sense of belonging.

Idea 1

💬 Group chat recommendations for students who have not joined a group

Idea 2

⚡️ Match students with other like-minded individuals

Idea 3

📱 Create a Forum where students can post questions for other students or university staff members to answer

Idea 4

🎙️ Broadcast area for universities to post user generated content

Rapid testing and experimentation

To explore and assess which idea could bring the most value we conducted various different activities to test hypothesis and help drive the decision on what to build first.

Student Matching

  • 💬 If we provide students with group recommendations that closely match their course and interests, we will see an increase in yield as students will feel engaged with the university and foster a sense of belonging with like-minded students

  • - Students are often invited to generic facebook groups which are irrelevant or so over-populated that they are nervous to post a message

    - Being in an online group makes students feel engaged and supported - Students want to get to know others on their course

  • In-app experiment: we asked students to select their main areas of interest and then used tags to generate personalised recommendations of what groups to join that exist in Community

  • 🔴 The current tagging system is determined by what tags universities use when creating group chats. This means the recommendation engine is not very sophisticated and in most cases, the tags did not present very interesting groups, thus resulting in no increase of engagement. To make this effective, we learnt that the groups need to be focused on areas of interest such as sports.

Chat Recommendations

  • ⚡️ If we match students with other like-minded individuals we will see an increase in yield as users will feel more of a sense of belonging to the university from a reduction in anxiety as a result to connecting with other individuals prior to enroling

  • - Students are mostly anxious about whether they will make friends in the lead up to enrolling. Some students have not enrolled because of this worry

    - Students want to know if there are others with similar interests that they could join clubs with

  • Offer holders were recruited via usertesting to interact with a low-fidelity student matching prototype so we could assess reactions and desirability. A series of questions were asked to give us an overall desirability rating.

    We also conducted interviews with universities to share the idea and assess desirability/feasibility.

  • 🟢 Students were very receptive to the idea and liked the concept. 70% of app users would like to be matched with other students in a direct message that share similar interests. usertesting results suggest students are excited by the idea as they would like to create some contacts before enrolling Universities were very fond of this idea as it required no effort on their side, it is automated through Community

Q&A Forum

  • 📱 If we provide students with a secure and private space to ask questions during the summer period, we will see an increase in yield as students will feel well-informed and confident in their decision from receiving quick responses to their worries

  • - Students have specific concerns which they need clarification on but find it takes a long time to receive an email back from the university; universities are time poor and low in resource so cannot reply to everyone quickly

  • Via user testing, offer holders were recruited via usertesting to interact with a low-fidelity prototype to assess reactions and desirability. Interviews with universities to share the idea and assess desirability/feasibility.

  • 🟠 Universities were apprehensive as it would mean they need to respond to more questions in another format which they do not have time for. But they like that all students would have visibility of the questions asked. Students were receptive but would want it to be anonymous just in case their questions seemed “embarrassing”

Broadcast Area

  • 🎙️ If we provide universities with a broadcast area we will see an increase in yield as students will feel more informed and engaged, contributing towards a greater sense of belonging before enrolling. This instills confidence in students having made the right decision

  • - Students find it hard to find content about the university and surrounding area which is not created for marketing purposes

    - Universities are unaware as to whether students have read the emails they send with important information

  • Discovery interviews with universities to assess whether this concept is feasible and how the experience could look. I created a low fidelity prototype to share with customers during the calls

  • 🟠 Universities liked the idea of having one single place to push content but feel that it would need a lot of time which they do not currently have. The optimum experience would be for Community to pull through content from across the web that relates to the university.

Student Matching experimentation

In-app survey

Through testing 4 different hypothesis, we identified Student Matching to be considered the most valuable for both offer holder students and universities.

We were able to identify the value of providing Student Matching to Unibuddy Community by conducting an in-app survey. This allowed us to assess interest and how users would prefer to be matched.

Findings

The Experiment results suggest students want to be matched with other incoming students that are on the same course, share similar interests and those who they might be living with. Some students would like to be randomly matched.


Qualitative research supports the idea that students do want to get to know other like-minded individuals before arriving on campus. This is because students’ no.1 worry before enrolling is whether they will form friendships. By connecting with other students before enrolling it can help to instil confidence and reduce this worry.

Hypothesis

Universities will see an increase in conversion if we can help students to foster a sense of belonging by alleviating the anxiety over whether they will form friendships.

We will achieve this by connecting students with other like-minded individuals within Community. This will be a success if we see 5% of all messages sent are between matched students.

MVP definition

Students are most concerned about whether they would fit in and enjoy their several years to come at university. Being able to connect and chat with like-minded students prior to enrolling would bring a lot of comfort, reassurance and confidence in their decision to enrol at University.

User story mapping and user flow definition

As a Product Trio we defined the MVP through a story mapping exercise which we then sliced down into several slices. The first slice focuses on collecting students’ subject area and then matching them with others who are studying the same subject. This is because we learnt that students want to connect with others who will be studying on the same course because they will spend the vast majority of their time with those individuals.

It was important to consider two different journeys for collecting subject information as new students will be onboarding to Unibuddy Community when this MVP is released but will also need to be accessible for those already existing as a user.

Mobile first

When defining the user experience for student matching, I worked very closely with the engineers to create an MVP which was accessible, and feasible. At the time, I was working with a very under developed design system so the success of delivery was underpinned by close collaboration with my engineering team.

In this scenario, it was very much about “right now vs right way” so we could assess the impact of student matching with our users.

MVP testing

5 students were recruited via usertesting for an unmoderated test where they were required to interact with the student matching feature. The aim of this test was to further validate that this concept would bring value to students before enrolling.


The test was a success and further validated that Student Matching could serve as an effective USP for the business whilst helping to solve the pain points encountered by students and university staff. Additional insights were gained in this test that surfaced further opportunities to enhance and optimise Student Matching in the future.

💡 All users felt Student Matching would help break the barriers to communication, particularly in a post-covid world where social interaction is quite anxiety inducing. They felt that this feature would be most useful in the before starting university and within the first year.

💡 It was also identified that it would be great for users to connect with others who shared similar interests/hobbies so they could create connections they can socialise with outside of the studying.

Final MVP

Student matching

Design decisions & delivery

📊 Data driven

We sliced the MVP into manageable development chunks. This was achieved by understanding the key tasks a user would need in order to be matched with another student and to engage - this was informed through user testing and discovery insights

♻️ Recycling of components

At the time, I was working with a very underdeveloped design system but through close collaboration with the developers we were able to create an MVP that took less time, effort and cost to build so we could quickly test our hypothesis

🤝 Close collaboration

Working in an agile way, I was able to sync up with developers and other areas of the business throughout the process to reach the final solution

Post-MVP optimisations

We quickly identified that whilst students were engaging with the matching feature, they were not initiating conversations with one another.

Ice breakers

A phenomena was identified which we called “first message anxiety” where students were matching but not conversing. We took a proactive approach and decided to implement ice breakers to encourage students to interact with one another.

The most common ice breaker used was “Hey! 👋 How are you?”

Push notifications

To inspire students to interact with the matching feature we also introduced push notifications. These notifications lead the user to view their matches.

⭐️ The total number of messages sent in matched conversations doubled the following day after push notifications went live.

MVP Outcomes

The first iteration of student matching proved a success and signalled that Student Matching is incredibly valued by students waiting to enrol. Universities were also fond of the feature and saw an increase in conversion.

+22%

increase in undergraduate conversion seen by University of Leicester

+33%

increase of all messages sent in Community were between matched students (this is 28% more than originally defined as our success metric for the hypothesis)

50%

of all direct messages were sent between matched students

Challenges and learnings

🧪 Testing & experimentation

I would like to have done a opportunity gap scoring exercise with each idea to help with prioritisation of future opportunities. It would also have been a more fair approach to defining the value of each idea.

As people, we tend to form connections over shared interests and whilst it’s useful for students to connect with others studying the same subject I do believe students would be more likely to engage with others over a longer period of time which would drive a more sustainable level of engagement as people adopt and change their hobbies over time. If I were to do this again, I’d like to have A/B tested the impact by also matching students according to their hobbies.

🤔 Deeper competitor analysis

I would also have liked for the product team to do a deeper competitor analysis (this is because I have learnt how to do it well at British Airways) so that we could identify future opportunities to stay ahead of the competition as it is a very competitive market.

⏰ Time & budget constraints

The main challenge was the timelines. We had to work so incredibly fast whilst managing other priorities with other squads. Therefore the outcome is a representation of right now, vs right way.

We had access to user testing tools but very limited credits so we couldn’t test as much as I would like to. For instance, I would like to have tested with users who have digital accessibility needs to check the feature was inclusive and accessible.